GOOGLE LLC patent applications published on December 28th, 2023

Patent applications for GOOGLE LLC on December 28th, 2023

REMOTE REAL-TIME GAME INPUT AGGREGATION (18035412)

Main Inventor

Collin Irwin

Brief explanation

The abstract describes a patent application for techniques and systems that generate and display aggregated gaming actions based on proposed game inputs provided by multiple remote client devices. These inputs are related to real-time display of gaming content that is based on the interactions of a first user with a remote gaming device.

• The system receives proposed game inputs from multiple remote audience users associated with remote client devices.
• These inputs are provided during the real-time display in response to a solicitation.
• An aggregated game action is generated based on the multiple proposed game inputs.
• The system then provides an indication of the aggregated game action to the remote gaming device.

Potential Applications

• Multiplayer gaming experiences with real-time interaction from remote audience users.
• Enhanced engagement and interactivity in live streaming of gaming content.
• Crowd-sourced decision-making in gaming scenarios.

Problems Solved

• Lack of real-time interaction and engagement from remote audience users in gaming experiences.
• Limited interactivity and decision-making options for viewers of live streaming gaming content.
• Difficulty in incorporating crowd-sourced inputs into gaming scenarios.

Benefits

• Increased engagement and interactivity for remote audience users in gaming experiences.
• Enhanced viewer experience in live streaming gaming content.
• More dynamic and varied gaming scenarios through crowd-sourced inputs.

Abstract

Techniques and systems are provided for generating and displaying aggregated gaming actions based on proposed game inputs provided via each of multiple remote client devices, related to real-time display of gaining content that is based at least in part on interactions of a first user with a remote gaming device. An indication of one or more proposed game inputs is received from multiple respective remote audience users associated with the multiple remote client devices during the real-time display in response to an indicated solicitation. An aggregated game action is generated based at least in part on the multiple proposed game inputs, and an indication of the aggregated game action is provided to the remote gaming device.

Updating Map Data With User-Generated Image Data (17782235)

Main Inventor

Subhasish Roy

Brief explanation

The technology described in this patent application is about updating map data in a mapping service. Here is a simplified explanation of the abstract:

• A mapping service can receive permission from a user to contribute image data to the map service.
• The image data captured by the user is uploaded to the map service by a backend service.
• The map service can analyze the image data and provide a recommendation to the backend service.
• The recommendation includes identifying points of interest in the map data that correspond to the image data.
• If no points of interest are found, the recommendation indicates that.
• The backend service can then request the selection of a point of interest if the recommendation includes one or more points.
• Alternatively, the backend service can request permission from the user to add a new point of interest to the map data.

Potential Applications

This technology has potential applications in various industries and services, including:

• Mapping and navigation services
• Travel and tourism applications
• Real estate and property management
• Local business directories
• Urban planning and development

Problems Solved

This technology solves several problems related to updating map data:

• Manual updating of map data can be time-consuming and inefficient.
• Identifying points of interest in image data can be challenging without automated assistance.
• Adding new points of interest to map data may require user permission and verification.

Benefits

The benefits of this technology include:

• Streamlining the process of updating map data by automating the identification of points of interest.
• Allowing users to contribute image data to improve the accuracy and completeness of map services.
• Enabling the addition of new points of interest based on user contributions, enhancing the overall map data.

Abstract

The technology relates to updating map data of a mapping service. A mapping service may receive permission to contribute image data captured by a user to the map service. The image data may be uploaded to the map service by a backend service. The map service may provide a recommendation to the backend service including an identification of points of interest in the map data that the image data corresponds to, or an indication that no points of interest were found within the map data that correspond with the image data. The backend service may request the selection of a points of interest the image corresponds to when the recommendation includes one or more points, or request permission from the user to add a new point of interest to the map data.

METHODS, SYSTEMS, AND MEDIA FOR AUTOMATIC AND CONTINUOUS CONTROL OF ENERGY-CONSUMING DEVICES (17848125)

Main Inventor

Charlotte Matthews

Brief explanation

The patent application describes methods, systems, and media for automatic and continuous control of energy-consuming devices. Here is a simplified explanation of the abstract:

• The method involves determining eligible devices for automatic control that are connected to a communications network.
• Available actions for each connected device are determined.
• Automatic control actions are generated for a portion of the connected devices based on the dynamic rate of electricity generated by a dynamic rate engine and the energy mode selected by a user in an application.
• A user interface is presented in an application to indicate the automatic control actions, with an option for the user to override them.
• If the user does not select the override option, a subset of the automatic control actions is transmitted to communication gateways connected to the devices, and another subset is transmitted directly to the devices via a device cloud platform.
• The user interface is updated to indicate the performed automatic control actions and the energy usage savings resulting from them.

Potential applications of this technology:

• Smart homes: The technology can be used to automatically control energy-consuming devices in a home, optimizing energy usage and reducing costs.
• Industrial automation: It can be applied in industrial settings to control and manage energy consumption of various devices and equipment.
• Energy management systems: The technology can be integrated into energy management systems to provide automated control and optimization of energy usage.

Problems solved by this technology:

• Manual control: The technology eliminates the need for manual control of energy-consuming devices by providing automatic control based on dynamic electricity rates and user preferences.
• Energy wastage: By continuously monitoring and controlling devices, the technology helps reduce energy wastage and improve energy efficiency.
• User convenience: The user interface allows users to easily view and override automatic control actions, providing flexibility and convenience.

Benefits of this technology:

• Energy savings: The automatic control actions based on dynamic electricity rates and user preferences can lead to significant energy savings.
• Cost reduction: By optimizing energy usage, the technology can help reduce electricity bills and overall energy costs.
• Improved efficiency: Continuous monitoring and control of energy-consuming devices improve overall energy efficiency and reduce wastage.

Abstract

Methods, systems, and media for automatic and continuous control of energy-consuming devices are provided. In some embodiments, the method comprises: determining, using a hardware processor, connected devices from a plurality of devices that are eligible for automatic control, wherein the plurality of devices are connected to a communications network; determining, using the hardware processor, available actions for each of the connected devices; generating, using the hardware processor, a plurality of automatic control actions for at least a portion of the connected devices based at least in part on a dynamic rate of electricity generated by a dynamic rate engine and an energy mode selected by a user in an application, wherein the plurality of automatic control actions are based on the available actions for each connected device; causing, using the hardware processor, a user interface to be presented in an application that indicates the plurality of automatic control actions for transmission to the portion of the connected devices, wherein the user interface provides a user interface element for overriding one or more of the plurality of automatic control actions; in response to determining that the user interface element for overriding one or more of the plurality of automatic control actions has not been selected by the user of the application, transmitting, using the hardware processor, a first subset of the plurality of automatic control actions to one or more communication gateways that are each connected to at least one of the subset of connected devices and a second subset of the plurality of automatic control actions directly to a subset of the connected devices via a device cloud platform; and updating, using the hardware processor, the user interface to indicate that at least one of the first subset of the plurality of automatic control actions and the second subset of the plurality of automatic control actions have been performed on the portion of the connected devices and energy usage savings from the performance of the automatic control actions.

Smart Context Subsampling On-Device System (18243338)

Main Inventor

Alexander Varshavsky

Brief explanation

The present disclosure is about a system that intelligently samples information on a device, such as location and activities, using machine learning to optimize battery usage and maintain or improve the quality of reported metrics. 

• The system intelligently samples information on a device, such as location and activities.
• Machine learning is used to optimize the sampling and uploading of background context.
• The optimization reduces battery usage while maintaining or improving the quality of reported metrics.
• A policy is generated based on the machine learning, dictating how scanning and upload rates should change in response to device conditions.

Potential Applications

This technology can be applied in various fields where intelligent sampling of information is required, such as:

• Location-based services: Enhancing the accuracy and efficiency of location tracking for navigation, delivery services, and geolocation-based applications.
• Activity tracking: Improving the monitoring and analysis of user activities for fitness tracking, healthcare, and productivity applications.
• Data analytics: Providing high-quality data for analysis and insights in fields like market research, urban planning, and transportation optimization.

Problems Solved

This technology addresses several problems related to sampling information on a device:

• Battery drain: By optimizing the sampling and uploading process, the system reduces battery usage, extending device battery life.
• Data quality: The machine learning-based approach ensures that the reported metrics maintain or improve their quality, providing more accurate and reliable information.
• Resource efficiency: The system intelligently adjusts scanning and upload rates based on device conditions, optimizing resource utilization and minimizing unnecessary data collection.

Benefits

The benefits of this technology include:

• Extended battery life: By reducing battery usage, users can enjoy longer device usage without the need for frequent charging.
• Improved data accuracy: The machine learning optimization enhances the quality of reported metrics, leading to more accurate and reliable information.
• Efficient resource utilization: The system intelligently adjusts sampling rates, optimizing resource usage and reducing unnecessary data collection, resulting in improved device performance.

Abstract

The present disclosure provides a system for intelligently sampling information, such as location, activities, etc. on device. Sampling and uploading of background context is optimized using machine learning, such that battery usage is reduced, and quality of metrics based on the reported information is maintained or improved. A policy is generated based on the machine learning, the policy dictating how scanning and upload rates should change in response to conditions on the device.

DEVICE TRACKING WITH INTEGRATED ANGLE-OF-ARRIVAL DATA (18035970)

Main Inventor

Steven Benjamin Goldberg

Brief explanation

The patent application describes a wearable device, such as a head-wearable display (HWD), that can determine its position using a combination of angle-of-arrival (AOA) data and inertial data. This data is fused together using techniques like stochastic estimation or machine learning to accurately identify the device's pose (position and orientation). The computer device associated with the HWD can then use this pose information to modify virtual reality or augmented reality content, providing an immersive experience for the user.

• The wearable device uses AOA data and inertial data to determine its position.
• Data integration techniques like stochastic estimation and machine learning are used to fuse the AOA and inertial data.
• The computer device associated with the wearable device uses the fused data to identify the device's pose.
• The identified pose is used to modify virtual reality or augmented reality content for an immersive user experience.

Potential Applications

• Virtual reality gaming and entertainment
• Augmented reality applications for training and education
• Industrial applications for remote assistance and visualization

Problems Solved

• Accurately determining the position and orientation of a wearable device
• Providing a seamless and immersive user experience in virtual reality or augmented reality environments

Benefits

• Improved accuracy in determining the pose of a wearable device
• Enhanced user experience in virtual reality or augmented reality content
• Increased potential for applications in various industries such as gaming, education, and remote assistance.

Abstract

A wearable device, such as a head-wearable display (HWD), identifies a device pose based on a combination of angle-of-arrival (AOA) data generated by, for example, an ultra-wideband (UWB) positioning module, and inertial data generated by an inertial measurement unit (IMU). The HWD fuses the AOA data with the inertial data using data integration techniques such as one or more of stochastic estimation (e.g., a Kalman filter), a machine learning model, and the like, or any combination thereof. A computer device associated with the HWD can employ the fused data to identify a pose of the HWD (e.g., a six degree of freedom (6 DoF) pose) and can use the identified pose to modify virtual reality or augmented reality content implemented by the computer device, thereby providing an immersive and enjoyable experience for a user.

TOUCH CONTROL OF WEARABLE DEVICES (18006096)

Main Inventor

Dongeek Shin

Brief explanation

The abstract describes a wearable device with a sensor that can detect and recognize gestures to control the device. The sensor, such as a PPG sensor on a smartwatch, can detect light reflected from underneath the device and from a touch to the wrist. The detected light is filtered to isolate the light from the touch, and a waterfall image is generated to analyze the changes in the isolated light over time and amplitude. This allows the device to detect and recognize gestures performed on the wrist, such as a touch, which can supplement the touch area provided by the device's display.

• Sensor on a wearable device can detect and recognize gestures to control the device
• Photoplethysmography (PPG) sensor on a smartwatch can detect light reflected from underneath the device and from a touch to the wrist
• Detected light is filtered to isolate the light from the touch
• Waterfall image is generated to analyze changes in the isolated light over time and amplitude
• Allows the device to detect and recognize gestures performed on the wrist, such as a touch
• Supplement the touch area provided by the device's display

Potential Applications

• Enhanced gesture control for wearable devices
• Improved user interaction with smartwatches and other wearable devices
• Potential applications in fitness tracking, healthcare monitoring, and virtual reality

Problems Solved

• Limited touch area on wearable devices
• Difficulty in controlling wearable devices with small displays
• Lack of intuitive and convenient gesture control options

Benefits

• Expanded touch area for controlling wearable devices
• Improved user experience and convenience
• Enhanced functionality and versatility of wearable devices

Abstract

A sensor on a wearable device may be further configured as a sensor for detecting and recognizing a gesture to control the wearable device. For example, light detected by a photoplethysmography (PPG) sensor of a smart watch may include (i) light back reflected from underneath the smart watch and (ii) light back reflected from a touch to a wrist adjacent to the smart watch. The detected light may be filtered to isolate the light back reflected from the touch. A waterfall image that includes information about how the isolated light changes with time and amplitude may be generated and used to detect and recognize gestures performed on the wrist, such as a touch. This additional touch area may help to supplement a touch area provided by a display to control the smart watch.

LIST NAVIGATION WITH EYE TRACKING (18340612)

Main Inventor

Erik Hubert Dolly Goossens

Brief explanation

The described techniques allow for displaying a group of positions on a screen, with some positions empty and others filled with a subset of list elements. The filled positions include an adjacent position next to the empty ones. 

• Movement of a gaze point can be tracked relative to the filled positions and around a central location towards the empty positions.
• The empty positions can be filled with list elements, and the adjacent position can be emptied to display a different subset of list elements.

Potential applications of this technology:

• User interfaces for touchless devices or eye-tracking systems.
• Displaying and navigating through lists or menus.
• Virtual reality or augmented reality experiences.

Problems solved by this technology:

• Facilitates efficient and intuitive navigation through lists or menus.
• Enables touchless interaction with devices, reducing the need for physical input methods.
• Enhances accessibility for individuals with limited mobility or dexterity.

Benefits of this technology:

• Improved user experience by providing a natural and intuitive way to navigate through content.
• Increased accessibility for individuals with disabilities.
• Enables new interaction possibilities in touchless or eye-tracking systems.

Abstract

Described techniques enable displaying a plurality of positions arranged around a central location on a display, the plurality of positions including at least one empty position and filled positions displaying a first subset of list elements, the filled positions including an adjacent position that is adjacent to the at least one empty position. Movement of a gaze point may be tracked, relative to the filled positions and around the central location towards the at least one empty position. The at least one empty position may be filled with a list element of the list elements, and the adjacent position may be emptied to thereby display a second subset of the list elements.

PREEMPTION IN A MACHINE LEARNING HARDWARE ACCELERATOR (18036506)

Main Inventor

Temitayo Fadelu

Brief explanation

The present disclosure describes a system and method for preempting a long-running process with a higher priority process in a machine learning system, such as a hardware accelerator.

• The system and method are applicable to machine learning hardware accelerators, which are multi-chip systems including semiconductor chips designed for machine learning operations.
• The hardware accelerator can include application-specific integrated circuits (ASICs), which are customized integrated circuits designed for specific uses.
• The innovation focuses on preempting a long-running process, which means interrupting and stopping a process that is taking a long time to complete.
• The preempting is done by introducing a higher priority process, which takes precedence over the long-running process and is given more resources and attention.
• This preempting mechanism allows for better resource allocation and prioritization in the machine learning system, leading to improved performance and efficiency.

Potential Applications

• This technology can be applied in various machine learning systems, particularly those that utilize hardware accelerators.
• It can be used in data centers and cloud computing environments where machine learning operations are performed.
• The technology can also be implemented in edge devices and IoT devices that require efficient machine learning capabilities.

Problems Solved

• Long-running processes in machine learning systems can hinder overall system performance and efficiency.
• Prioritizing and allocating resources to different processes can be challenging in complex machine learning systems.
• This technology solves these problems by preempting long-running processes and introducing higher priority processes, improving resource allocation and system performance.

Benefits

• Improved performance and efficiency in machine learning systems.
• Better resource allocation and prioritization of processes.
• Enhanced responsiveness and reduced latency in executing machine learning operations.

Abstract

The present disclosure describes a system and method for preempting a long-running process with a higher priority process in a machine learning system, such as a hardware accelerator. The machine learning hardware accelerator can be a multi-chip system including semiconductor chips that can be application-specific integrated circuits (ASIC) designed to perform machine learning operations. An ASIC is an integrated circuit (IC) that is customized for a particular use.

Generating Deeplinks for Applications Based on Multi-Level Referrer Data (18243385)

Main Inventor

Justin Lewis

Brief explanation

The abstract describes a method in a computing device that involves generating a set of data packets containing referrer data associated with resources rendered by the device. These resources are related to different applications. The data packets are then transmitted to an application server, which generates a deeplink or referrer tag based on the referrer data. The second application on the computing device retrieves content associated with the referrer data using the deeplink or referrer tag, and displays it within its own content interface.

• Method involves generating data packets with referrer data for different resources rendered by a computing device.
• The referrer data includes information about the resources and the applications associated with them.
• Data packets are sent to an application server, which generates a deeplink or referrer tag based on the referrer data.
• The second application on the computing device retrieves content using the deeplink or referrer tag.
• The content is then displayed within the content interface of the second application.

Potential Applications

• This technology can be used in mobile apps to provide seamless integration between different applications.
• It can be utilized in e-commerce apps to track and analyze user behavior across different resources and applications.
• Content sharing platforms can use this technology to enable users to easily share and access content from different sources.

Problems Solved

• Simplifies the process of sharing and accessing content across different applications on a computing device.
• Provides a standardized method for generating deeplinks or referrer tags based on referrer data.
• Enables seamless integration and communication between different applications.

Benefits

• Enhances user experience by allowing easy navigation between different applications and resources.
• Enables personalized content recommendations based on user behavior across different applications.
• Facilitates data analysis and tracking of user interactions across different resources and applications.

Abstract

A method in a computing device includes generating a set of data packets comprising a first referrer data associated with a first resource rendered by the computing device when executing a first application and a second referrer data associated with a second resource rendered by the computing device prior to rendering the first resource. The first resource provides a first content item associated with a second application. The method also includes transmitting the set of data packets to an application server, receiving, from the application server, a deeplink, or a referrer tag, that was generated based on the first referrer data and the second referrer data, and retrieving, by the second application and using the deeplink or the referrer tag, content associated with the first referrer data and the second referrer data for display at the computing device within a content interface of the second application.

Controlling Memory Frequency Based on Transaction Queue Occupancy (18253174)

Main Inventor

Derek James Basehore

Brief explanation

The patent application describes techniques and apparatuses that use transaction queue lengths to adjust the clock frequency controlling memory access in an electronic device.

• Detecting when a transaction queue threshold is exceeded.
• Initiating a counter to measure the duration of the violation.
• Determining if the transaction queue threshold continues to be violated for the measured duration.
• Altering the clock frequency that controls memory access in response to the violation.

Potential Applications

This technology can be applied in various electronic devices that utilize memory access, such as:

• Computers and servers.
• Mobile devices (smartphones, tablets).
• Internet of Things (IoT) devices.
• Embedded systems.

Problems Solved

The technology addresses the following problems:

• Transaction queue congestion leading to performance degradation.
• Inefficient memory access due to high queue lengths.
• Bottlenecks in electronic devices caused by memory access delays.

Benefits

The use of transaction queue lengths to adjust clock frequency offers several benefits:

• Improved performance by dynamically adapting the clock frequency based on queue length.
• Efficient memory access management, reducing congestion and delays.
• Enhanced overall system responsiveness and throughput.
• Optimal utilization of memory resources in electronic devices.

Abstract

Techniques and apparatuses are described that use transaction queue lengths to alter a clock frequency that controls access to a memory of an electronic device. Techniques include detecting that a transaction queue threshold has been violated, initiating a counter to measure a time duration, determining that the transaction queue threshold continues to be violated for the time duration and, in response, altering the clock frequency that controls access to the memory of the electronic device.

In-Memory Distributed Cache (18464068)

Main Inventor

Asa Briggs

Brief explanation

The abstract describes a method for an in-memory distributed cache that allows a client device to write data to the random access memory (RAM) of a memory host. Here is a simplified explanation of the abstract:

• The method receives a write request from a client device to write a block of client data in the RAM of a memory host.
• It determines whether the client device has permission to write the block of client data at the memory host.
• It also checks if the block of client data is already saved at the memory host and if a free block of RAM is available.
• If the client device has permission, the block of client data is not already saved, and a free block of RAM is available, the write request is allowed.
• The client is then allowed to write the block of client data to the free block of RAM.

Potential applications of this technology:

• Distributed databases: This method can be used in distributed databases to improve performance by caching frequently accessed data in memory.
• Content delivery networks (CDNs): CDNs can utilize this method to cache popular content closer to end-users, reducing latency and improving content delivery speed.
• Real-time analytics: In-memory caching can be beneficial for real-time analytics platforms, allowing faster data processing and analysis.

Problems solved by this technology:

• Performance optimization: By caching data in memory, the method reduces the need to retrieve data from slower storage devices, improving overall system performance.
• Scalability: The distributed nature of the cache allows for scaling across multiple memory hosts, accommodating larger amounts of data and increasing system capacity.

Benefits of this technology:

• Faster data access: Storing data in memory enables quicker retrieval and processing, leading to reduced latency and improved response times.
• Improved system performance: By reducing the reliance on slower storage devices, the method enhances the overall performance of the system.
• Scalability and flexibility: The distributed cache can be easily scaled by adding more memory hosts, allowing for increased storage capacity and accommodating growing data demands.

Abstract

A method for an in-memory distributed cache includes receiving a write request from a client device to write a block of client data in random access memory (RAM) of a memory host and determining whether to allow the write request by determining whether the client device has permission to write the block of client data at the memory host, determining whether the block of client data is currently saved at the memory host, and determining whether a free block of RAM is available. When the client device has permission to write the block of client data at the memory host, the block of client data is not currently saved at the memory host, and a free block of RAM is available, the write request is allowed and the client is allowed to write the block of client data to the free block of RAM.

Heterogeneous Compute Platform Architecture For Efficient Hosting Of Network Functions (18213028)

Main Inventor

Santanu Dasgupta

Brief explanation

The present disclosure describes a converged compute platform architecture that includes two configurations: an infrastructure processing unit (IPU)-only configuration and a configuration where the IPU is coupled with a central processing unit (CPU) like an x86 processor. 

• The first configuration consists of only the IPU, while the second configuration combines the IPU with a CPU.
• The two configurations can communicate with each other through a PCIe switch or remote direct memory access (RDMA) techniques.
• Both configurations utilize machine learning (ML) acceleration through a single converged architecture.

Potential Applications

This technology has potential applications in various fields, including:

• Data centers and cloud computing environments
• Artificial intelligence and machine learning applications
• High-performance computing and scientific research

Problems Solved

The converged compute platform architecture addresses the following problems:

• Efficient utilization of resources by combining the IPU and CPU in a single architecture
• Improved communication and connectivity between the two configurations
• Accelerated machine learning capabilities through a converged architecture

Benefits

The benefits of this technology include:

• Enhanced performance and efficiency in data centers and cloud computing environments
• Improved ML acceleration through a converged architecture
• Simplified communication and connectivity between the IPU and CPU configurations

Abstract

The present disclosure provides for a converged compute platform architecture, including a first infrastructure processing unit (IPU)-only configuration and a second configuration wherein the IPU is coupled to a central processing unit, such as an x86 processor. Connectivity between the two configurations may be accomplished with a PCIe switch, or the two configurations may communicate through remote direct memory access (RDMA) techniques. Both configurations may use ML acceleration through a single converged architecture.

APPROXIMATE K NEAREST NEIGHBORS ON HARDWARE ACCELERATORS (18341697)

Main Inventor

Felix Ren-Chyan Chern

Brief explanation

The patent application describes methods, systems, and apparatus for performing a kNN computation using a hardware accelerator. The abstract outlines the steps involved in the process, including obtaining query vectors and database vectors, and using a hardware accelerator to compute similarity values and identify the most similar database vectors for each query vector.

• The patent application focuses on performing a kNN computation using a hardware accelerator.
• The method involves obtaining a set of query vectors and a set of database vectors.
• The hardware accelerator is used to compute similarity values between each query vector and each database vector.
• For each query vector, the hardware accelerator identifies the index and similarity value of the most similar database vector within each bin.

Potential Applications

• This technology can be applied in various fields that require similarity-based computations, such as machine learning, data mining, and recommendation systems.
• It can be used to improve the efficiency and speed of kNN computations in large-scale datasets.

Problems Solved

• Traditional kNN computations can be computationally expensive and time-consuming, especially for large datasets.
• This technology solves the problem by utilizing a hardware accelerator to perform the computations, resulting in faster and more efficient processing.

Benefits

• The use of a hardware accelerator improves the speed and efficiency of kNN computations.
• It allows for faster processing of large-scale datasets, enabling real-time or near-real-time applications.
• The technology can be integrated into existing systems and workflows, enhancing their performance without significant modifications.

Abstract

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for performing a kNN computation using a hardware accelerator. One of the methods includes obtaining a set of one or more query vectors; obtaining a set of database vectors; and performing, on a hardware accelerator and for each query vector in the set, a search for the k most similar database vectors to the query vector, comprising: computing, by circuitry of the hardware accelerator and for each query vector, a respective similarity value between the query vector and each database vector; and for each query vector, identifying, by the hardware accelerator and for each bin, (i) an index of the most similar database vector within the bin and (ii) the respective similarity value for the most similar database vector within the bin.

METHODS, SYSTEMS, AND MEDIA FOR PRESENTING CONTENT BASED ON A GENERIC RATING (18244483)

Main Inventor

Joon-Hee Jeon

Brief explanation

The patent application describes methods, systems, and media for presenting content based on a generic rating. It involves converting country-specific content ratings to generic ratings and blocking search results based on user-selected content rating restrictions. The modified search results are then presented to the user. When the user selects content, its country-specific rating is converted to a generic rating and checked against the user's content rating restriction. If it is not blocked, the selected content is presented.

• The patent application presents a method for presenting content based on a generic rating.
• It involves converting country-specific content ratings to generic ratings.
• The method allows for blocking search results based on user-selected content rating restrictions.
• Modified search results are presented to the user after removing blocked search results.
• The method also converts the country-specific content rating of selected content to a generic rating.
• It checks if the selected content is blocked based on the generic rating and user's content rating restriction.
• If not blocked, the selected content is presented to the user.

Potential Applications

This technology can have various applications, including:

• Content filtering and censorship in search engines and online platforms.
• Customized content presentation based on user preferences and restrictions.
• Compliance with content rating regulations across different countries.
• Enhancing user experience by filtering out unwanted or inappropriate content.

Problems Solved

The technology addresses several problems, such as:

• Ensuring compliance with content rating regulations in different countries.
• Providing users with control over the type of content they can access.
• Filtering out inappropriate or unwanted content from search results.
• Simplifying the process of converting and comparing content ratings across different rating systems.

Benefits

The technology offers several benefits, including:

• Improved content filtering and censorship capabilities.
• Enhanced user control and customization of content presentation.
• Simplified conversion and comparison of content ratings.
• Compliance with content rating regulations in different countries.
• Enhanced user experience by filtering out unwanted or inappropriate content.

Abstract

Methods, systems, and media for presenting content based on a generic rating are provided. In some implementations, the method comprises: receiving search results; determining country-specific content ratings associated with the search results; converting the country-specific content ratings to generic content ratings associated with the search results; determining that at least one search result is to be blocked based on the generic content ratings and a user-selected generic content rating restriction; in response to determining that a search result is to be blocked, removing the search result from the search results to create modified search results; causing the modified search results to be presented; receiving a selection of content from the presented search results; determining a country-specific content rating associated with the selected content; converting the country-specific content rating to a generic content rating; determining that the selected content is not to be blocked based on the generic content rating and the user-selected generic content rating restriction; and causing the selected content to be presented.

Systems and Methods of Person Recognition in Video Streams (18462986)

Main Inventor

Akshay R. Bapat

Brief explanation

The abstract describes a method for recognizing persons in video streams and handling their identification. Here is a simplified explanation of the abstract:

• The method starts by obtaining a live video stream.
• It then detects the presence of a person in the video stream.
• Through analysis of the video stream, the method determines certain information about the person's attributes.
• Based on this information, the method determines if the person is identifiable to the computing system or not.
• If the person is not identifiable, the method stores some of the information obtained.
• The method also allows for user input to classify the person as a stranger.
• Finally, if the person is classified as a stranger, the method deletes the stored information.

Potential applications of this technology:

• Video surveillance systems: The method can be used in security systems to identify and handle unknown individuals captured in video streams.
• Privacy protection: By deleting stored information of unidentified individuals, the method helps protect privacy rights.

Problems solved by this technology:

• Identification of unknown individuals: The method provides a way to handle the identification of people who are not identifiable to the system, such as strangers.
• Privacy concerns: By deleting stored information, the method addresses privacy concerns related to the collection and storage of personal data.

Benefits of this technology:

• Enhanced security: The method allows for the identification and handling of unknown individuals, potentially improving security measures.
• Privacy preservation: By deleting stored information, the method respects privacy rights and helps mitigate privacy risks associated with video surveillance systems.

Abstract

A method for recognizing persons in video streams includes obtaining a live video stream, detecting a first person in the live video stream, determining from analysis of the live video stream first information that identifies an attribute of the first person, determining based on at least some of the first information that the first person is not identifiable to the computing system, storing at least some of the first information, receiving a user classification of the first person as being a stranger, and deleting the stored first information.

Machine Learning Based Document Visual Element Extraction (17808293)

Main Inventor

Nikolay Glushnev

Brief explanation

==Abstract Explanation==

The patent application describes a method that involves analyzing a document containing both text and visual elements. The method uses machine learning models to determine the location of each textual field and the visual element within the document. It then assigns a visual element anchor token to the visual element and inserts it into the textual fields based on its location and the location of the textual fields. After inserting the visual element anchor token, the method extracts structured entities representing the textual fields and the visual element using a text-based extraction model.

• The method analyzes a document with both text and visual elements.
• It determines the location of each textual field and the visual element within the document.
• It assigns a visual element anchor token to the visual element.
• The visual element anchor token is inserted into the textual fields based on their respective locations.
• The method extracts structured entities representing the textual fields and the visual element using a text-based extraction model.

Potential Applications

• Document analysis and organization
• Data extraction from documents with mixed text and visual elements
• Content management systems
• Information retrieval and indexing

Problems Solved

• Efficiently analyzing documents with both text and visual elements
• Accurately determining the location of textual fields and visual elements within a document
• Extracting structured entities from documents with mixed content

Benefits

• Improved document analysis and organization
• Enhanced data extraction capabilities
• Streamlined content management processes
• More efficient information retrieval and indexing

Abstract

A method includes obtaining a document with textual fields and a visual element. For each textual field, the method includes determining a textual offset for the textual field that indicates a location of the textual field relative to each other textual field in the document. The method includes detecting, using a machine learning vision model, the visual element and determining a visual element offset indicating a location of the visual element relative to each textual field in the document. The method includes assigning the visual element a visual element anchor token and inserting the visual element anchor token into the textual fields in an order based on the visual element offset and the respective textual offsets. The method also includes, after inserting the visual element anchor token, extracting, using a text-based extraction model, from the textual fields, structured entities representing the series of textual fields and the visual element.

Systems And Methods For Training Translation Models Using Source-Augmented Training Examples (17988315)

Main Inventor

Jing Huang

Brief explanation

The patent application describes systems and methods for training a translation model using text sequences in different languages and a label based on the source of the text sequence.

• The translation model is trained using a first text sequence in one language, a second text sequence in a different language, and a label indicating the source of the second text sequence.
• The label can include information such as an Internet domain, subdomain, URL, website name, or IP address.
• The label may also indicate the source of the first text sequence.
• Training examples are automatically generated by sampling the first text sequence from a page of an Internet domain, sampling the second text sequence from another page of the same domain, and generating the label based on the source data of the second page.

Potential applications of this technology:

• Improving machine translation systems by training them on specific domains or sources.
• Enhancing translation accuracy by incorporating information about the source of the text.
• Enabling more targeted translation models for specific websites or IP addresses.

Problems solved by this technology:

• Overcoming limitations of generic translation models by training them on specific domains or sources.
• Addressing the challenge of accurately translating text from different languages by considering the source of the text.
• Providing a more customized and accurate translation experience for specific websites or IP addresses.

Benefits of this technology:

• Improved translation quality and accuracy.
• Enhanced understanding of the context and source of the text being translated.
• Customized translation models for specific domains or sources, leading to more relevant and precise translations.

Abstract

Systems and methods for training a translation model based on a first text sequence in a first language, a second text sequence in a second language different from the first language, and a label based on a source of the second text sequence. In some examples, the label may comprise an Internet domain, an Internet subdomain, a uniform resource locator, a website name, or an IP address. In some examples, the label may further indicate a source of the first text sequence. In some examples, each given training example may be automatically generated by sampling the first text sequence from a first page of a given Internet domain, sampling the second text sequence from a second page of the given Internet domain, and generating the label based on all or a portion of source data of the second page.

MIXTURE OF EXPERTS NEURAL NETWORKS (18244171)

Main Inventor

Noam M. Shazeer

Brief explanation

The patent application describes a system that includes a neural network with a Mixture of Experts (MoE) subnetwork. The MoE subnetwork consists of multiple expert neural networks that process the output of the first neural network layer.

• The MoE subnetwork includes expert neural networks that process the first layer output to generate expert outputs.
• A gating subsystem selects one or more expert neural networks based on the first layer output and assigns a weight to each selected expert neural network.
• The first layer output is provided as input to each selected expert neural network.
• The expert outputs generated by the selected expert neural networks are combined according to their weights to generate an MoE output.
• The MoE output is then used as input to the second neural network layer.

Potential Applications

• This technology can be applied in various fields where complex data processing is required, such as image recognition, natural language processing, and speech recognition.
• It can be used in autonomous vehicles for tasks like object detection and classification.
• It can be utilized in recommendation systems to provide personalized recommendations based on user preferences.

Problems Solved

• The Mixture of Experts (MoE) subnetwork allows for more efficient and accurate processing of complex data by utilizing multiple expert neural networks.
• It addresses the challenge of handling diverse and complex data by combining the outputs of different expert neural networks.
• The gating subsystem helps in selecting the most relevant expert neural networks based on the input, improving the overall performance of the system.

Benefits

• The system improves the accuracy and efficiency of data processing by leveraging the strengths of multiple expert neural networks.
• It allows for better handling of complex and diverse data by combining the outputs of different expert neural networks.
• The gating subsystem enhances the adaptability of the system by dynamically selecting the most suitable expert neural networks based on the input.

Abstract

A system includes a neural network that includes a Mixture of Experts (MoE) subnetwork between a first neural network layer and a second neural network layer. The MoE subnetwork includes multiple expert neural networks. Each expert neural network is configured to process a first layer output generated by the first neural network layer to generate a respective expert output. The MoE subnetwork further includes a gating subsystem that selects, based on the first layer output, one or more of the expert neural networks and determine a respective weight for each selected expert neural network, provides the first layer output as input to each of the selected expert neural networks, combines the expert outputs generated by the selected expert neural networks in accordance with the weights for the selected expert neural networks to generate an MoE output, and provides the MoE output as input to the second neural network layer.

Improved Processing of Sequential Data via Machine Learning Models Featuring Temporal Residual Connections (18253416)

Main Inventor

Liangzhe Yuan

Brief explanation

The patent application describes a system and method that uses a machine-learned model, such as a convolutional neural network, with temporal residual connections. These connections allow the model to transfer intermediate feature data from one instantiation of the model to another, enabling the processing of sequential inputs.

• The machine-learned model includes one or more temporal residual connections.
• Each temporal residual connection supplies intermediate feature data from the current instantiation of the model to other instantiations of the model.
• The other instantiations can process subsequent or preceding sequential inputs.
• This technology can be applied to various sequential input processing tasks.

Potential Applications

• Image recognition and classification
• Natural language processing
• Speech recognition
• Video analysis and processing

Problems Solved

• Efficient processing of sequential inputs
• Improved accuracy in sequential input analysis
• Handling of temporal dependencies in data

Benefits

• Enhanced performance in sequential input processing
• Better utilization of intermediate feature data
• Improved accuracy and efficiency in machine learning models

Abstract

Systems and methods can include or leverage a machine-learned model (e.g., a convolutional neural network) that includes one or more temporal residual connections. In particular, each temporal residual connection can respectively supply one or more sets of intermediate feature data generated by a current instantiation of the model from a current sequential input to one or more other instantiations of the machine-learned model applied to process one or more other sequential inputs. For example, the other instantiations of the machine-learned model can include subsequent instantiations of the machine-learned model applied to process one or more subsequent sequential inputs that follow the current sequential input in a sequence and/or preceding instantiations of the machine-learned model applied to process one or more preceding sequential inputs that precede the current sequential input in a sequence.

Adaptive Learning Rates for Training Adversarial Models with Improved Computational Efficiency (18341397)

Main Inventor

Hussein Hazimeh

Brief explanation

The patent application describes systems and methods for dynamically adjusting the learning rate of an adversarial model using a novel scheduling technique. The technique aims to maintain a balance between the adversarial components of the model. Here are the key points:

• The learning rate of an adversarial model is adjusted dynamically using a unique scheduling technique.
• The technique is based on the fact that the loss of an ideal adversarial network can be determined in advance in certain scenarios.
• A scheduler component is employed to ensure that the loss of the optimized network remains close to that of an ideal adversarial network.

Potential Applications

• This technology can be applied in various fields where adversarial models are used, such as computer vision, natural language processing, and cybersecurity.
• It can enhance the performance and stability of adversarial models in tasks like image classification, text generation, and anomaly detection.

Problems Solved

• Adversarial models often struggle to maintain a proper balance between their components, leading to suboptimal performance.
• Determining an appropriate learning rate for adversarial models can be challenging, as it needs to be adjusted dynamically based on the specific scenario.
• This technology addresses these issues by providing a novel scheduling technique that adapts the learning rate to maintain the desired balance and improve overall performance.

Benefits

• The dynamic learning rate scheduling technique ensures that the adversarial model remains optimized and performs closer to an ideal adversarial network.
• By maintaining a proper balance between adversarial components, the model can achieve better accuracy and stability.
• The technique simplifies the process of determining an appropriate learning rate for adversarial models, making them more accessible and easier to implement.

Abstract

Provided are systems and methods that use a novel learning rate scheduling technique to dynamically adapt the learning rate of an adversarial model to maintain an appropriate balance between adversarial components of the model. The scheduling technique is driven by the fact that, in some settings, the loss of an ideal adversarial network can be analytically determined a priori. A scheduler component can thus operate to keep the loss of the optimized network close to that of an ideal adversarial net.

SYSTEMS AND METHODS FOR DETERMINING THAT A MEDIA ITEM IS BEING PRESENTED (18244511)

Main Inventor

Vincent Dureau

Brief explanation

The patent application describes systems and methods for determining if media items are currently being presented to a user. Here are the key points:

• The method involves identifying a media item that may be presented on an output device connected to a client device.
• A level of confidence is calculated to determine if the identified media item is currently being presented to the user.
• After calculating the level of confidence, the method determines if any predetermined events associated with user interaction with the media item have occurred.
• Based on these predetermined events, the level of confidence that the identified media item is being presented to the user is increased or decreased.

Potential applications of this technology:

• Media monitoring: This technology can be used to track and monitor media items being presented to users, such as advertisements or promotional content.
• Content recommendation: By determining which media items are being presented to users, this technology can help in recommending relevant content based on user preferences and interactions.
• User engagement analysis: The method can be used to analyze user engagement with media items by tracking the occurrence of predetermined events during their presentation.

Problems solved by this technology:

• Accurate media item detection: The method provides a way to accurately identify if a media item is currently being presented to a user, even in complex scenarios where multiple media items may be simultaneously presented.
• User interaction tracking: By tracking predetermined events associated with user interaction, the method enables a more comprehensive analysis of user engagement with media items.
• Confidence level adjustment: The ability to dynamically adjust the level of confidence based on user interactions allows for more accurate determination of whether a media item is being presented.

Benefits of this technology:

• Improved media monitoring: The method provides a more accurate and automated way to monitor the presentation of media items, enabling better tracking and analysis of advertising campaigns or content distribution.
• Enhanced user experience: By accurately determining which media items are being presented, this technology can help in delivering personalized and relevant content to users, improving their overall experience.
• Data-driven insights: The method allows for detailed analysis of user engagement with media items, providing valuable insights for content creators, advertisers, and marketers.

Abstract

The various implementations described herein include systems and methods for determining that media items are currently being presented. In one aspect, a method performed at a client device includes: (1) identifying a media item potentially being presented on an output device coupled to the client device; (2) calculating a level of confidence that the identified media item is currently being presented to the user; (3) subsequent to the calculating, determining that at least one predetermined event associated with user interaction with the media item has occurred; and (4) based on the at least one predetermined event, increasing or decreasing the level of confidence that the identified media item is currently being presented to the user.

Identifying Consumers in a Transaction Via Facial Recognition (18322197)

Main Inventor

Sashikanth Chandrasekaran

Brief explanation

The patent application describes a system for processing payments using facial recognition technology. Here is a simplified explanation of the abstract:

• A merchant and a user register with a payment processing system.
• The user's facial image is used to create a facial template.
• The user signs into a payment application on their device.
• The device receives an identifier from a merchant beacon device.
• The identifier is transmitted to the payment processing system.
• The payment processing system sends facial templates to the merchant camera device for other users in range of the beacon.
• The merchant camera device compares captured facial images with the received templates to identify the user.
• The merchant selects the user's account on their POS device.
• Transaction details are sent to the payment processing system.
• The payment processing system processes the transaction with an issuer system.
• The payment processing system receives approval for the transaction and sends a receipt to the merchant POS device.

Potential Applications

• Secure and convenient payment processing in various retail settings.
• Streamlining the payment process by eliminating the need for physical cards or devices.
• Enhancing customer experience by providing a seamless and personalized payment experience.

Problems Solved

• Reduces the risk of fraud by using facial recognition for user identification.
• Simplifies the payment process by eliminating the need for physical cards or devices.
• Improves security by using facial templates for user authentication.

Benefits

• Increased security through facial recognition technology.
• Improved convenience for users by eliminating the need for physical payment methods.
• Streamlined payment process for merchants, reducing transaction time and improving efficiency.

Abstract

A merchant and a user register with a payment processing system, which establishes a facial template based on a user image. The user signs into a payment application via a user computing device, which receives an identifier from a merchant beacon device to transmit to the payment processing system. The payment processing system transmits facial templates to the merchant camera device for other users who are also signed in to the payment application in range of the merchant beacon device. The merchant camera device compares a captured facial image against the received facial templates to identify the user. A merchant POS device operator selects an account of the user. The merchant POS device transmits transaction details to the payment processing system, which processes the transaction with an issuer system. The payment processing system receives an approval of the transaction authorization request and transmits a receipt to the merchant POS device.

APPLICATION PROGRAM INTERFACE SCRIPT CACHING AND BATCHING (18232559)

Main Inventor

Varouj A. Chitilian

Brief explanation

==Abstract==

The patent application describes methods, systems, and apparatus for managing application program interface (API) calls. It also includes computer programs encoded on a computer storage medium for implementing these methods and systems.

Patent/Innovation Explanation

• The patent application focuses on managing API calls.
• It includes methods, systems, and apparatus for this purpose.
• Computer programs encoded on a computer storage medium are provided to implement these methods and systems.

Potential Applications

This technology can be applied in various industries and scenarios, including:

• Software development and programming
• Web development and API integration
• Cloud computing and distributed systems
• Internet of Things (IoT) applications
• Mobile app development

Problems Solved

The technology addresses several problems related to managing API calls, such as:

• Ensuring efficient and reliable communication between different software components
• Handling and processing large volumes of API calls
• Managing API authentication and security
• Monitoring and analyzing API usage and performance
• Facilitating integration and interoperability between different software systems

Benefits

The benefits of this technology include:

• Improved efficiency and reliability of API communication
• Enhanced scalability and performance for handling large volumes of API calls
• Simplified management of API authentication and security measures
• Better monitoring and analysis of API usage and performance metrics
• Streamlined integration and interoperability between software systems

Abstract

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for managing application program interface calls.

UNSUPERVISED DEPTH PREDICTION NEURAL NETWORKS (18367888)

Main Inventor

Vincent Michael Casser

Brief explanation

The patent application describes a system for generating a depth output for an image. Here are the key points:

• The system takes input images of the same scene, each with potential objects.
• It generates a background image for each input image and analyzes them to determine camera motion.
• For each potential object, it generates an object motion output based on the input images and camera motion.
• The system uses a depth prediction neural network to process a particular input image and generate a depth output.
• The parameters of the depth prediction neural network are updated based on the depth output, camera motion, and object motion.

Potential applications of this technology:

• Computer vision systems that require accurate depth information for image processing.
• Augmented reality applications that need to understand the depth of objects in the real world.
• Autonomous vehicles that rely on depth perception for navigation and obstacle avoidance.

Problems solved by this technology:

• Accurately determining the depth of objects in an image can be challenging, especially when there is camera motion and potential objects in the scene.
• This system addresses these challenges by considering camera motion and object motion to generate a more accurate depth output.

Benefits of this technology:

• Improved accuracy in generating depth outputs for images.
• The ability to update the parameters of the depth prediction neural network based on the specific scene and camera/object motion, leading to better performance over time.
• The system can be applied to various applications that require depth information, enhancing their capabilities.

Abstract

A system for generating a depth output for an image is described. The system receives input images that depict the same scene, each input image including one or more potential objects. The system generates, for each input image, a respective background image and processes the background images to generate a camera motion output that characterizes the motion of the camera between the input images. For each potential object, the system generates a respective object motion output for the potential object based on the input images and the camera motion output. The system processes a particular input image of the input images using a depth prediction neural network (NN) to generate a depth output for the particular input image, and updates the current values of parameters of the depth prediction NN based on the particular depth output, the camera motion output, and the object motion outputs for the potential objects.

Pose Empowered RGB-Flow Net (18464912)

Main Inventor

Yinxiao Li

Brief explanation

The patent application describes a method for analyzing video data of an actor performing an activity using neural networks. 

• The video data is processed to generate three input streams: spatial images representing spatial features of the actor, temporal images representing motion, and pose images representing the actor's pose.
• These input streams are then processed by at least one neural network.
• The neural network classifies the activity based on the information from the spatial, temporal, and pose input streams.

Potential applications of this technology:

• Video surveillance systems that can automatically detect and classify activities performed by individuals.
• Sports analysis tools that can analyze the movements and poses of athletes to provide insights and feedback.
• Virtual reality and augmented reality applications that can track and analyze the movements of users for interactive experiences.

Problems solved by this technology:

• Automating the analysis of video data to classify activities, reducing the need for manual review and analysis.
• Providing a more comprehensive understanding of an actor's activity by considering spatial, temporal, and pose information together.
• Enabling real-time analysis and classification of activities, allowing for immediate response or feedback.

Benefits of this technology:

• Improved accuracy and efficiency in classifying activities compared to traditional methods.
• Ability to analyze and understand complex activities that involve both spatial and temporal aspects.
• Potential for a wide range of applications in various industries, including security, sports, and entertainment.

Abstract

A method includes receiving video data that includes a series of frames of image data. Here, the video data is representative of an actor performing an activity. The method also includes processing the video data to generate a spatial input stream including a series of spatial images representative of spatial features of the actor performing the activity, a temporal input stream representative of motion of the actor performing the activity, and a pose input stream including a series of images representative of a pose of the actor performing the activity. Using at least one neural network, the method also includes processing the temporal input stream, the spatial input stream, and the pose input stream. The method also includes classifying, by the at least one neural network, the activity based on the temporal input stream, the spatial input stream, and the pose input stream.

CHANNEL-WISE AUTOREGRESSIVE ENTROPY MODELS FOR IMAGE COMPRESSION (18461292)

Main Inventor

David Charles Minnen

Brief explanation

The patent application describes methods, systems, and apparatus for channel-wise autoregressive entropy models. These models are used to process data and generate compressed representations of the data. Here are the key points:

• The method involves using a first encoder neural network to generate a latent representation of the data.
• The latent representation is then processed by a quantizer and a second encoder neural network to generate a quantized latent representation of the data and a latent representation of an entropy model.
• The quantized latent representation is further processed into multiple slices of quantized latent representations, arranged in an ordinal sequence.
• A hyperprior processing network generates hyperprior parameters and a compressed representation of these parameters.
• For each slice of the quantized latent representations, a corresponding compressed representation is generated using a slice processing network.
• The combination of these compressed representations forms a compressed representation of the data.

Potential applications of this technology:

• Data compression: The methods described can be used to compress data, reducing its size while preserving important information.
• Image and video compression: This technology can be applied to compress images and videos, enabling efficient storage and transmission.
• Data transmission: Compressed representations can be transmitted over networks more quickly and with less bandwidth usage.

Problems solved by this technology:

• Efficient compression: The methods described provide a way to compress data effectively, reducing storage and transmission requirements.
• Preserving important information: The techniques used ensure that important information is retained in the compressed representation.

Benefits of this technology:

• Reduced storage requirements: Compressed representations take up less space, allowing for more efficient storage of large amounts of data.
• Faster data transmission: Compressed representations can be transmitted more quickly over networks, reducing latency and improving overall performance.
• Preserved data quality: Despite compression, the important information in the data is retained, ensuring minimal loss of quality.

Abstract

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for channel-wise autoregressive entropy models. In one aspect, a method includes processing data using a first encoder neural network to generate a latent representation of the data. The latent representation of data is processed by a quantizer and a second encoder neural network to generate a quantized latent representation of data and a latent representation of an entropy model. The latent representation of data is further processed into a plurality of slices of quantized latent representations of data wherein the slices are arranged in an ordinal sequence. A hyperprior processing network generates a hyperprior parameters and a compressed representation of the hyperprior parameters. For each slice, a corresponding compressed representation is generated using a corresponding slice processing network wherein a combination of the compressed representations form a compressed representation of the data.

Geocoding Personal Information (18202619)

Main Inventor

Adam Bliss

Brief explanation

The abstract describes a computer-based method for generating geocoded user information. Here is a simplified explanation of the abstract:

• The method involves searching user data from various sources for entries that contain location-related information.
• The locations for the identified entries are determined.
• A map is generated that displays the current location of a mobile device and representations of the entries with location-related information at their respective locations.
• The entries can come from multiple different data corpuses, meaning various sources of user data.

Potential applications of this technology:

• Location-based advertising: The generated map can be used to display targeted advertisements based on the user's location and the entries with location-related information.
• Social networking: The method can be used to show the locations of friends or contacts on a map, enhancing social networking experiences.
• Travel planning: The generated map can assist users in planning trips by displaying relevant information about various locations.

Problems solved by this technology:

• Efficient data retrieval: The method allows for searching and retrieving user data from multiple sources, making it easier to gather location-related information.
• Geocoding accuracy: By determining the locations for the entries, the method ensures accurate representation of the data on the map.

Benefits of this technology:

• Enhanced user experience: The generated map provides a visual representation of the user's location and relevant entries, improving the overall user experience.
• Personalized information: By displaying entries with location-related information, the method offers personalized and contextually relevant data to the user.
• Efficient data organization: The method allows for organizing and presenting user data in a geographically meaningful way, making it easier to understand and analyze.

Abstract

A computer-implemented method for generating geocoded user information is disclosed. The method comprises searching user data across multiple different data corpuses for entries having location-related information and determining locations for the location-related information. The method further comprises generating a map showing a current location of a mobile device along with representations of the entries having location-related information, at the determined locations, for entries from the multiple different data corpuses.

VOLUMETRIC PERFORMANCE CAPTURE WITH NEURAL RENDERING (18251743)

Main Inventor

Sean Ryan Francesco FANELLO

Brief explanation

The patent application describes a technique for capturing the performance of a subject in a three-dimensional (volumetric) manner using neural rendering. Here are the key points:

• The technique involves capturing images of the subject from multiple viewpoints and lighting conditions using a light stage, along with depth data obtained from infrared cameras.
• A neural network is used to extract features of the subject from the images based on the depth data.
• These features are then mapped into a texture space (UV texture space) for further processing.
• A neural renderer is employed to generate an output image of the subject from a desired viewpoint, ensuring that the illumination in the output image aligns with the target view.
• The neural renderer achieves this by resampling the features of the subject from the texture space to an image space, resulting in the generation of the output image.

Potential applications of this technology:

• Film and entertainment industry: This technique can be used to capture performances of actors or characters in a more realistic and immersive manner, enabling the creation of high-quality visual effects.
• Virtual reality (VR) and augmented reality (AR): By capturing volumetric performances, this technique can enhance the realism and interactivity of VR and AR experiences.
• Gaming: Game developers can utilize this technique to capture and render realistic character animations, improving the overall gaming experience.

Problems solved by this technology:

• Traditional methods of capturing performances often rely on markers or sensors attached to the subject, which can be cumbersome and limit the freedom of movement. This technique eliminates the need for physical markers or sensors.
• The use of neural rendering allows for more accurate and realistic rendering of the subject, including proper alignment of illumination with the target view.

Benefits of this technology:

• Improved realism: The use of neural rendering and volumetric capture techniques enhances the realism of captured performances, resulting in more immersive and engaging visual content.
• Increased efficiency: By eliminating the need for physical markers or sensors, the capture process becomes more efficient and less intrusive for the subject.
• Flexibility and freedom of movement: The technique allows for capturing performances with natural movement and without restrictions, enabling more authentic and dynamic representations.

Abstract

Example embodiments relate to techniques for volumetric performance capture with neural rendering. A technique may involve initially obtaining images that depict a subject from multiple viewpoints and under various lighting conditions using a light stage and depth data corresponding to the subject using infrared cameras. A neural network may extract features of the subject from the images based on the depth data and map the features into a texture space (e.g., the UV texture space). A neural renderer can be used to generate an output image depicting the subject from a target view such that illumination of the subject in the output image aligns with the target view. The neural render may resample the features of the subject from the texture space to an image space to generate the output image.

Fingerprint Enrollment Using Collocation of a User's Touch and a Fingerprint Sensor (18250464)

Main Inventor

Firas Sammoura

Brief explanation

This patent application describes a system for safeguarding a computing device using a fingerprint identification system that utilizes biometric data. The system includes a fingerprint sensor that is used during the enrollment process to collect the user's fingerprint data, such as from their thumb, finger, palm, etc. The system guides the user to complete the enrollment process by combining their touch (e.g., a thumb-tap) with the location of the fingerprint sensor, making it easier and requiring fewer thumb-taps. This improves the user experience while maintaining biometric security.

• The patent application describes a system for safeguarding a computing device using a fingerprint identification system.
• The system includes a fingerprint sensor used during the enrollment process to collect the user's biometric data.
• The biometric data can include fingerprint data from various parts of the user's hand, such as the thumb, finger, palm, etc.
• The system guides the user to complete the enrollment process by combining their touch (e.g., a thumb-tap) with the location of the fingerprint sensor.
• This combination of touch and sensor location makes the enrollment process easier and requires fewer thumb-taps.
• The techniques described in the patent application improve the user experience while maintaining biometric security.

Potential Applications

• This technology can be applied to any computing device that requires user authentication, such as smartphones, tablets, laptops, etc.
• It can be used in various industries where secure access to devices is crucial, such as banking, healthcare, government, etc.
• This technology can also be integrated into physical access control systems, providing secure entry to buildings or restricted areas.

Problems Solved

• Simplifies the enrollment process for fingerprint identification systems, making it more user-friendly.
• Reduces the number of thumb-taps required during the enrollment process, improving efficiency.
• Enhances the overall user experience by providing guidance and ease of use.
• Maintains biometric security while improving the enrollment process.

Benefits

• Improved user experience: The combination of touch and sensor location makes the enrollment process easier and more intuitive for users.
• Enhanced security: Biometric data, such as fingerprints, provides a high level of security for device authentication.
• Efficiency: The system reduces the number of thumb-taps required during the enrollment process, saving time and effort for users.
• Versatility: The system can collect biometric data from various parts of the hand, allowing for flexibility in user authentication.

Abstract

This disclosure describes apparatuses, methods, and techniques for enabling a user to safeguard a computing device with a fingerprint identification system by using biometric data. The fingerprint identification system includes a fingerprint sensor used during an enrollment process of the user's biometric data. The biometric data may include fingerprint data from the user's thumb, finger, a plurality of fingers, palm, and so forth. The computing device uses a collocation of a user's touch, for example, a thumb-tap, and a fingerprint sensor's location to guide the user to complete the enrollment process of a full fingerprint with ease and with fewer thumb-taps. Consequently, the techniques enable biometric security with an enrollment process having a good user experience.

SIMULTANEOUS ACOUSTIC EVENT DETECTION ACROSS MULTIPLE ASSISTANT DEVICES (18367859)

Main Inventor

Matthew Sharifi

Brief explanation

The patent application describes a system that can detect and process audio data captured by multiple assistant devices in an ecosystem. Here are the key points:

• The system can detect audio data capturing an acoustic event at multiple assistant devices.
• Each assistant device processes the audio data locally to generate measures associated with the acoustic event using event detection models.
• The measures are then processed to determine if the detected acoustic event is an actual event.
• If the event is determined to be actual, an action associated with the event is performed in response.

Potential applications of this technology:

• Smart home systems: The system can be used to detect and respond to various acoustic events in a smart home, such as a doorbell ringing or a smoke alarm going off.
• Voice assistants: The technology can enhance the capabilities of voice assistants by allowing them to detect and respond to specific acoustic events, such as a user saying a wake word or a specific command.
• Security systems: The system can be integrated into security systems to detect and respond to suspicious acoustic events, such as glass breaking or an alarm sounding.

Problems solved by this technology:

• Improved accuracy: By processing audio data locally at each assistant device, the system can reduce latency and improve the accuracy of detecting and responding to acoustic events.
• Privacy concerns: Processing audio data locally ensures that sensitive information is not sent to a central server, addressing privacy concerns associated with audio data collection.

Benefits of this technology:

• Faster response time: By processing audio data locally, the system can quickly detect and respond to acoustic events, providing a faster and more seamless user experience.
• Enhanced user privacy: Processing audio data locally reduces the need for sending sensitive information to a central server, ensuring user privacy.
• Scalability: The system can be deployed across multiple assistant devices in an ecosystem, allowing for scalability and widespread adoption.

Abstract

Implementations can detect respective audio data that captures an acoustic event at multiple assistant devices in an ecosystem that includes a plurality of assistant devices, process the respective audio data locally at each of the multiple assistant devices to generate respective measures that are associated with the acoustic event using respective event detection models, process the respective measures to determine whether the detected acoustic event is an actual acoustic event, and cause an action associated with the actional acoustic event to be performed in response to determining that the detected acoustic event is the actual acoustic event. In some implementations, the multiple assistant devices that detected the respective audio data are anticipated to detect the respective audio data that captures the actual acoustic event based on a plurality of historical acoustic events being detected at each of the multiple assistant devices.

DETERMINING STATE OF AUTOMATED ASSISTANT DIALOG (18367785)

Main Inventor

Abhinav Rastogi

Brief explanation

The abstract describes a patent application that focuses on determining the state of a conversation between an automated assistant and a user, and taking actions based on that state. The conversation state is represented by slots, each with candidate values and corresponding scores. The candidate values are determined through language processing of user and system utterances. The scores for candidate values are generated using a memory network and a scoring model.

• The patent application aims to determine the state of a conversation between an automated assistant and a user.
• The conversation state is represented by slots, each with candidate values and scores.
• Candidate values for slots are determined through language processing of user and system utterances.
• The scores for candidate values are generated using a memory network and a scoring model.

Potential Applications

• Conversational AI systems
• Virtual assistants
• Customer service chatbots

Problems Solved

• Difficulty in determining the state of a conversation between an automated assistant and a user
• Challenges in accurately assigning scores to candidate values for conversation slots

Benefits

• Improved accuracy in determining conversation state
• Enhanced performance of automated assistants
• More efficient and effective customer service interactions

Abstract

Determining a dialog state of an electronic dialog that includes an automated assistant and at least one user, and performing action(s) based on the determined dialog state. The dialog state can be represented as one or more slots and, for each of the slots, one or more candidate values for the slot and a corresponding score (e.g., a probability) for each of the candidate values. Candidate values for a slot can be determined based on language processing of user utterance(s) and/or system utterance(s) during the dialog. In generating scores for candidate value(s) of a given slot at a given turn of an electronic dialog, various features are determined based on processing of the user utterance and the system utterance using a memory network. The various generated features can be processed using a scoring model to generate scores for candidate value(s) of the given slot at the given turn.

SELECTIVELY PROVIDING ENHANCED CLARIFICATION PROMPTS IN AUTOMATED ASSISTANT INTERACTIONS (18244762)

Main Inventor

Matthew Sharifi

Brief explanation

Implementations described in this patent application aim to improve the accuracy of voice recognition systems by addressing ambiguous spoken utterances. When an utterance can be interpreted as requesting multiple actions, an enhanced clarification prompt is provided to the user to disambiguate between the possible actions. This prompt goes beyond natural language and includes additional user interface elements to gather input.

• Implementations receive audio data capturing a spoken utterance.
• Based on the audio data, a recognition corresponding to the utterance is generated.
• If the recognition is ambiguous, meaning it can be interpreted as requesting different actions exclusively, an enhanced clarification prompt is provided.
• The enhanced clarification prompt includes elements beyond natural language to solicit further user input for disambiguation.
• This approach replaces the traditional natural language-only clarification prompt.

Potential applications of this technology:

• Voice assistants: Improving the accuracy and user experience of voice-controlled virtual assistants like Siri, Alexa, or Google Assistant.
• Automotive systems: Enhancing voice recognition systems in cars for safer and more efficient hands-free operation.
• Smart home devices: Improving voice-controlled devices like smart speakers or thermostats to better understand user commands.

Problems solved by this technology:

• Ambiguity in spoken utterances: Addressing situations where a voice command can have multiple exclusive interpretations, leading to incorrect actions or confusion.
• User frustration: Reducing frustration caused by voice assistants misunderstanding or misinterpreting commands.
• Efficiency: Streamlining the user interaction process by quickly resolving ambiguous requests.

Benefits of this technology:

• Improved accuracy: By providing an enhanced clarification prompt, the system can gather additional input to disambiguate user requests accurately.
• Enhanced user experience: The inclusion of non-verbal elements in the prompt makes it easier for users to clarify their intended actions.
• Time-saving: By quickly resolving ambiguous requests, the system can provide the desired response or action without unnecessary delays.

Abstract

Implementations described herein receive audio data that captures a spoken utterance, generate, based on processing the audio data, a recognition that corresponds to the spoken utterance, and determine, based on processing the recognition, that the spoken utterance is ambiguous (i.e., is interpretable as requesting performance of a first particular action exclusively and is also interpretable a second particular action exclusively). In response to determining that the spoken utterance is ambiguous, implementations determine to provide an enhanced clarification prompt that renders output that is in addition to natural language. The enhanced clarification prompt solicits further user interface input for disambiguating between the first particular action and the second particular action. Determining to provide the enhanced clarification prompt includes a current or prior determination to provide the enhanced clarification prompt instead of a natural language (NL) only clarification prompt that is restricted to rendering natural language.

RESOLVING UNIQUE PERSONAL IDENTIFIERS DURING CORRESPONDING CONVERSATIONS BETWEEN A VOICE BOT AND A HUMAN (18462787)

Main Inventor

Rafael Goldfarb

Brief explanation

Implementations of this patent application involve a voice bot using machine learning to identify unique personal identifiers during conversations with humans. These identifiers can be a sequence of alphanumeric characters specific to each individual. Here are the key points:

• The voice bot analyzes spoken utterances to identify unique personal identifiers.
• It generates candidate identifiers based on the speech hypotheses.
• The bot prompts the human for clarification on specific alphanumeric characters until it accurately predicts the unique personal identifier.
• Once identified, the bot can use these identifiers to perform further actions.
• The technology utilizes multiple layers of machine learning to improve accuracy and efficiency.

Potential applications of this technology:

• Customer service: Voice bots can use unique personal identifiers to provide personalized assistance to customers.
• Security: Identifying unique personal identifiers can help verify the identity of individuals during conversations.
• Personalized recommendations: By recognizing unique personal identifiers, voice bots can offer tailored recommendations based on individual preferences.

Problems solved by this technology:

• Efficient identification: The technology streamlines the process of identifying unique personal identifiers during conversations, reducing the need for manual input.
• Accuracy: By utilizing machine learning, the voice bot can accurately predict the correct unique personal identifier.
• Personalization: The technology enables voice bots to provide personalized experiences by recognizing individual identifiers.

Benefits of this technology:

• Improved customer experience: Voice bots can offer personalized assistance and recommendations, enhancing the overall customer experience.
• Enhanced security: By verifying unique personal identifiers, the technology helps prevent unauthorized access to sensitive information.
• Time-saving: The automated identification process saves time for both the voice bot and the human user.

Abstract

Implementations are directed to causing a voice bot to utilize a plurality of ML layers in resolving unique personal identifier(s) for a human while the voice bot is engaged in a corresponding conversation with the human. The unique personal identifier(s) can include a unique sequence of alphanumeric characters that is personal to the human. In some implementations, ASR speech hypothes(es) corresponding to spoken utterance(s) that include the unique personal identifier(s) can be processed to generate candidate unique personal identifier(s), given alphanumeric character(s) of the candidate unique personal identifier(s) can be selected, and the voice bot can prompt the human with clarification request(s) to clarify the given alphanumeric character(s) until it is predicted to correspond to the an actual unique personal identifier(s) for the human(s). The unique personal identifier(s) can then be utilized in performance of further action(s) by the voice bot and/or other systems.

Conditioned Separation of Arbitrary Sounds based on Machine Learning Models (17808653)

Main Inventor

Beat Gfeller

Brief explanation

The patent application describes methods for training a neural network to separate audio sources from a given audio waveform using both audio clips and textual descriptions of the audio as training data. The methods involve generating a shared representation of the audio and text, where the audio embedding and text embedding of a given audio clip are close to each other. This shared representation is then used to train a neural network to separate the target audio source from the input audio waveform.

• The patent application proposes a method for training a neural network to separate audio sources from an audio waveform.
• The method uses both audio clips and textual descriptions of the audio as training data.
• A shared representation is generated by embedding the audio and text, ensuring that the embeddings of a given audio clip and its textual description are close to each other.
• The shared representation is then used to train a neural network to separate the target audio source from the input audio waveform.

Potential Applications

• Speech enhancement: The technology can be used to separate speech from background noise in audio recordings, improving speech intelligibility.
• Music source separation: It can be applied to separate individual instruments or vocals from a music recording, allowing for remixing or isolating specific elements.
• Audio transcription: The technology can assist in transcribing audio recordings by separating different speakers or audio sources, making it easier to transcribe each source separately.

Problems Solved

• Difficulty in separating specific audio sources from a mixture of sounds in an audio waveform.
• Lack of training data that combines both audio clips and textual descriptions, making it challenging to train neural networks for audio source separation.

Benefits

• Improved accuracy: By incorporating textual descriptions, the neural network can better understand the target audio source, leading to more accurate separation.
• Versatility: The technology can be applied to various audio source separation tasks, such as speech enhancement, music source separation, and audio transcription.
• Enhanced user experience: Separating audio sources can improve the quality and intelligibility of audio recordings, benefiting users in different domains such as media production, transcription services, and communication systems.

Abstract

Example methods include receiving training data comprising a plurality of audio clips and a plurality of textual descriptions of audio. The methods include generating a shared representation comprising a joint embedding. An audio embedding of a given audio clip is within a threshold distance of a text embedding of a textual description of the given audio clip. The methods include generating, based on the joint embedding, a conditioning vector and training, based on the conditioning vector, a neural network to: receive (i) an input audio waveform, and (ii) an input comprising one or more of an input textual description of a target audio source in the input audio waveform, or an audio sample of the target audio source, separate audio corresponding to the target audio source from the input audio waveform, and output the separated audio corresponding to the target audio source in response to the receiving of the input.

Automatic Non-Linear Editing Style Transfer (18251838)

Main Inventor

Nathan Frey

Brief explanation

The present disclosure is about a system, method, and computer program for automated non-linear editing style transfer in videos. This technology aims to enhance the editing process by automatically adjusting content segments based on the identified shot boundaries, analyzed content, and determined editing style.

• The method involves identifying shot boundaries in a video.
• The identified content in each shot is analyzed using object detection techniques.
• The motion across frames within each shot is measured to determine the editing style.
• A set of target content is analyzed in relation to the identified content and editing style of a shot.
• A content segment is automatically adjusted from the set of target content based on the determined editing style of the shot.

Potential applications of this technology:

• Video editing software and platforms can utilize this technology to automate the process of applying different editing styles to video content.
• Content creators and filmmakers can save time and effort by automating the editing process and achieving desired visual effects.
• This technology can be used in video production for creating trailers, promotional videos, and other visually appealing content.

Problems solved by this technology:

• Manual editing can be time-consuming and requires expertise. This technology automates the editing process, reducing the need for manual intervention.
• Determining the appropriate editing style for each shot can be subjective and challenging. This technology uses motion analysis to objectively determine the editing style.
• Matching the content segment with the editing style can be a complex task. This technology analyzes the content and adjusts it accordingly, ensuring a seamless transition.

Benefits of this technology:

• Automation of the editing process saves time and effort for content creators.
• Consistent and professional editing styles can be achieved across different shots in a video.
• Objectively determining the editing style based on motion analysis improves the overall quality of the edited video.

Abstract

The present disclosure provides systems, methods, and computer program products for performing automated non-linear editing style transfer. A computer-implemented method may include determining one or more shot boundaries in a video, analyzing identified content in each of one or more shots in the video based on performing object detection, determining an editing style for each of the one or more shots in the video based at least in part on measuring motion across frames within the respective shots, determining a content segment to adjust from a set of target content based on analyzing the set of target content in view of the identified content and the determined editing style of a shot from the video, and automatically adjusting the content segment from the set of target content based at least in part on modifying the content segment with the determined editing style of the shot from the video.

DEEP TRENCH CAPACITORS EMBEDDED IN PACKAGE SUBSTRATE (18244716)

Main Inventor

Nam Hoon Kim

Brief explanation

The abstract describes a patent application related to deep trench capacitors embedded in a package substrate for an integrated circuit. The chip package includes an integrated circuit die with a power distribution circuit and a substrate with one or more cavities. Deep trench capacitors are placed in these cavities and connected to the power distribution circuit.

• The patent application is about deep trench capacitors embedded in a package substrate for an integrated circuit.
• The chip package includes an integrated circuit die with a power distribution circuit.
• The substrate used in the chip package is different from the integrated circuit and has one or more cavities.
• The cavities can be formed on either the first surface or the second surface of the substrate.
• One or more deep trench capacitors are placed in these cavities.
• The deep trench capacitors are connected to the power distribution circuit using conductors.

Potential Applications

• This technology can be applied in various electronic devices that use integrated circuits, such as smartphones, tablets, and computers.
• It can be used in power management systems to improve the efficiency and stability of power distribution in integrated circuits.
• The deep trench capacitors can be utilized in high-performance computing systems to enhance the overall performance and reliability.

Problems Solved

• The technology solves the problem of power distribution in integrated circuits by providing deep trench capacitors embedded in the package substrate.
• It addresses the need for improved power management and stability in electronic devices.
• The use of deep trench capacitors helps to reduce noise and improve signal integrity in integrated circuits.

Benefits

• The integration of deep trench capacitors in the package substrate simplifies the design and manufacturing process of integrated circuits.
• It improves the power distribution efficiency and stability, leading to better overall performance of electronic devices.
• The use of deep trench capacitors helps to reduce noise and improve the reliability of integrated circuits.

Abstract

This disclosure relates to deep trench capacitors embedded in a package substrate on which an integrated circuit is mounted. In some aspects, a chip package includes an integrated circuit die that has a power distribution circuit for one or more circuits of the integrated circuit. The chip package also includes a substrate different from the integrated circuit and having a first surface on which the integrated circuit die is mounted and a second surface opposite the first surface. The substrate includes one or more cavities formed in at least one of the first surface or the second surface. The chip package also includes one or more deep trench capacitors disposed in at least one of the one or more cavities. Each deep trench capacitor is connected to the power distribution circuit by conductors.

GENERATING SEQUENCES OF NETWORK DATA WHILE PREVENTING ACQUISITION OR MANIPULATION OF TIME DATA (18361303)

Main Inventor

Gang Wang

Brief explanation

The abstract describes a method for determining network measurements using encrypted impression and conversion data from multiple client devices. The method involves two aggregation servers that perform a multi-party computation process to generate chronological sequences of encrypted data and decrypt the encrypted data.

• The method involves receiving encrypted impression data from multiple client devices.
• A second aggregation server receives encrypted conversion data from at least a portion of the client devices.
• The first and second aggregation servers perform a multi-party computation process.
• The process generates chronological sequences of encrypted impression and conversion data.
• The encrypted data is then decrypted by the aggregation servers.

Potential Applications

This technology has potential applications in various fields, including:

• Advertising: The method can be used to analyze the effectiveness of advertising campaigns by tracking impressions and conversions from multiple client devices.
• Market Research: It can be utilized to gather data on consumer behavior and preferences by analyzing encrypted impression and conversion data.
• Network Optimization: The method can help optimize network performance by analyzing network measurements obtained from multiple client devices.

Problems Solved

The method addresses the following problems:

• Privacy Concerns: By using encrypted data, the method ensures the privacy of user information while still allowing for analysis and computation.
• Data Aggregation: The method efficiently aggregates impression and conversion data from multiple client devices, allowing for comprehensive analysis.
• Secure Computation: The multi-party computation process ensures secure and accurate decryption of the encrypted data.

Benefits

The use of this technology offers several benefits:

• Privacy Protection: User data remains encrypted throughout the process, ensuring the privacy and security of sensitive information.
• Comprehensive Analysis: The method allows for the analysis of both impression and conversion data, providing a more complete understanding of user behavior.
• Efficient Data Processing: The multi-party computation process enables efficient decryption and analysis of the encrypted data, saving time and resources.

Abstract

Methods, systems, and apparatus, including a method for determining network measurements. In some aspects, a method includes receiving, by a first aggregation server and from each of multiple client devices, encrypted impression data. A second aggregation server receives, from each of at least a portion of the multiple client devices, encrypted conversion data. The first aggregation server and the second aggregation server perform a multi-party computation process to generate chronological sequences of encrypted impression data and encrypted conversion data and to decrypt the encrypted impression data and the encrypted conversion data.

Secure Multi-Party Reach and Frequency Estimation (18334035)

Main Inventor

Craig Wright

Brief explanation

The patent application describes systems and methods for generating min-increment counting bloom filters in a networking environment. Here is a simplified explanation of the abstract:

• The system maintains a set of data records that include device identifiers and attributes associated with devices in a network.
• It generates a vector with coordinates corresponding to counter registers.
• It identifies hash functions to update a counting bloom filter.
• It hashes the data records to extract index values pointing to a set of counter registers.
• It increments the positions in the min-increment counting bloom filter corresponding to the minimum values of the counter registers.
• It obtains an aggregated public key and encrypts the counter registers using the aggregated shared key to generate an encrypted vector.
• The encrypted vector is transmitted to a networked worker computing device.

Potential applications of this technology:

• Network traffic analysis: The system can be used to analyze the count and frequency of device identifiers and attributes in a network, providing insights into network traffic patterns.
• Security monitoring: By tracking the count and frequency of device identifiers, the system can help detect and prevent unauthorized access or suspicious activities in a network.
• Resource allocation: The system can assist in optimizing resource allocation by providing information on the usage and demand of devices in a network.

Problems solved by this technology:

• Efficient counting: The min-increment counting bloom filter allows for efficient counting of device identifiers and attributes, reducing the computational overhead compared to traditional counting methods.
• Privacy protection: The encryption of the counter registers ensures that sensitive information about the devices in the network is protected during transmission.
• Scalability: The system can handle large amounts of data records and efficiently update the counting bloom filter, making it suitable for use in large-scale networking environments.

Benefits of this technology:

• Improved network analysis: The system provides accurate and real-time information on the count and frequency of device identifiers and attributes, enabling better network analysis and decision-making.
• Enhanced security: By monitoring device identifiers, the system can help identify potential security threats and take proactive measures to mitigate them.
• Privacy-preserving: The encryption of the counter registers ensures that sensitive information remains confidential, protecting the privacy of the devices and their users.

Abstract

Systems and methods for generating min-increment counting bloom filters to determine count and frequency of device identifiers and attributes in a networking environment are disclosed. The system can maintain a set of data records including device identifiers and attributes associated with device in a network. The system can generate a vector comprising coordinates corresponding to counter registers. The system can identify hash functions to update a counting bloom filter. The system can hash the data records to extract index values pointing to a set of counter registers. The system can increment the positions in the min-increment counting bloom filter corresponding to the minimum values of the counter registers. The system can obtain an aggregated public key comprising a public key. The system can encrypt the counter registers using the aggregated shared key to generate an encrypted vector. The system can transmit the encrypted vector to a networked worker computing device.

Adaptive Sounding Reference Signal Mapping for Improved Channel Estimation (18253014)

Main Inventor

Ming Sun

Brief explanation

The present disclosure describes a method for improving channel estimation in a user equipment (UE) by implementing adaptive sounding reference signal (SRS) mapping. This method involves generating a set of SRS symbols, including first and second SRS symbols. The offset for the second SRS symbol is determined based on the difference between the first and second radio chains of the UE. The first and second SRS symbols are then mapped to an antenna port of the first radio chain. The UE transmits the first SRS symbol to a base station via the antenna port of the first radio chain and transmits the second SRS symbol to the base station via the same antenna port while applying the offset to the first radio chain. This improves channel estimation for both uplink and downlink communications.

• A set of sounding reference signal (SRS) symbols are generated, including first and second SRS symbols.
• The offset for the second SRS symbol is determined based on the difference between the first and second radio chains of the UE.
• The first and second SRS symbols are mapped to an antenna port of the first radio chain.
• The first SRS symbol is transmitted to a base station via the antenna port of the first radio chain.
• The second SRS symbol is transmitted to the base station via the same antenna port while applying the offset to the first radio chain.
• This method improves channel estimation for both uplink and downlink communications.

Potential Applications

• This technology can be applied in wireless communication systems to improve channel estimation for both uplink and downlink communications.

Problems Solved

• The method solves the problem of inaccurate channel estimation in wireless communication systems.

Benefits

• Improved channel estimation leads to better overall performance and reliability of wireless communication systems.
• This method can be implemented in existing user equipment without requiring significant hardware changes.

Abstract

The present disclosure describes various aspects of adaptive sounding reference signal mapping that a user equipment (UE) implements to improve channel estimation. In aspects, a set of sounding reference signal (SRS) symbols are generated that include at least first and second SRS symbols. An offset for the second SRS symbol is determined based on a difference between a first radio chain and a second radio chain of the UE. The first and second SRS symbols are then mapped to an antenna port of the first radio chain. The UE transmits the first SRS symbol to a base station via the antenna port of the first radio chain and transmits the second SRS symbol to the base station via the antenna port of the first radio chain while the offset is applied to the first radio chain. By so doing, the UE may improve channel estimation for uplink and/or downlink communications.

PREVENTING FRAUD IN AGGREGATED NETWORK MEASUREMENTS (18341353)

Main Inventor

Gang Wang

Brief explanation

The patent application describes a method for preventing fraud by receiving measurement data from multiple client devices. Each client device uses a threshold encryption scheme to generate a group key and a group member key. When a threshold number of measurement data elements with the same group identifier are received, the network data is decrypted using the group member keys.

• The method prevents fraud by securely encrypting and decrypting network data.
• Multiple client devices can contribute to the encryption and decryption process.
• The threshold encryption scheme ensures that a certain number of measurement data elements are required for decryption, adding an extra layer of security.

Potential Applications

• This technology can be applied in digital advertising to prevent fraudulent activities such as fake impressions or conversions.
• It can be used in online transactions to ensure secure communication and prevent data breaches.
• The method can be implemented in various industries where secure data transmission is crucial, such as healthcare or finance.

Problems Solved

• Fraud prevention: The method addresses the problem of fraudulent activities by securely encrypting and decrypting network data, ensuring the authenticity of the measurement data.
• Data security: By using a threshold encryption scheme and group keys, the method provides a secure way to transmit and protect sensitive data.

Benefits

• Enhanced security: The method offers a more secure way to encrypt and decrypt network data, reducing the risk of fraud and unauthorized access.
• Improved accuracy: By requiring a threshold number of measurement data elements for decryption, the method ensures that only valid and reliable data is used for analysis.
• Scalability: The method can be applied to multiple client devices, allowing for a scalable and efficient fraud prevention system.

Abstract

Methods, systems, and apparatus, including a method for preventing fraud. In some aspects, a method includes: receiving, from multiple client devices, a measurement data element that includes a respective group member key and a group identifier for a given conversion as a result of displaying a digital component. Each client device uses a threshold encryption scheme to generate, based at least on network data that includes one or more of impression data or conversion data for the conversion, a group key that defines a secret for encrypting the network data and generate, based on data related to the application, the respective group member key that includes a respective share of the secret. In response to determining that at least the threshold number of measurement data elements having the same group identifier have been received, the network data is decrypted using the group member keys in the received measurement data elements.

Reliable Transport Protocol and Hardware Architecture for Datacenter Networking (18367679)

Main Inventor

Weihuang Wang

Brief explanation

==Abstract Explanation==

The abstract describes a communication protocol system that ensures reliable transmission of packets. It introduces an initiator entity that determines the need to send data to a target entity. The initiator entity sends a solicited push request to the target entity, asking it to receive the outgoing data. Upon receiving the request, the target entity responds with a push grant. Finally, the initiator entity sends the outgoing data to the target entity as requested.

Patent/Innovation Explanation

• The system provides a communication protocol for reliable packet transport.
• An initiator entity determines the need to transmit data to a target entity.
• The initiator entity sends a solicited push request to the target entity, asking it to receive the data.
• The target entity responds with a push grant, indicating its readiness to receive the data.
• The initiator entity then sends the outgoing data to the target entity.

Potential Applications

• This technology can be applied in various communication systems, such as network protocols, internet protocols, or wireless communication protocols.
• It can be used in data transmission between devices, ensuring reliable delivery of packets.

Problems Solved

• The system solves the problem of reliable packet transport by providing a communication protocol that ensures successful transmission.
• It addresses the issue of data loss or incomplete transmission by using a solicited push request and push grant mechanism.

Benefits

• The system ensures reliable transport of packets, reducing the risk of data loss or incomplete transmission.
• It provides a more efficient and effective communication protocol for transmitting data between entities.
• The use of solicited push requests and push grants improves the overall reliability and success rate of data transmission.

Abstract

A communication protocol system is provided for reliable transport of packets. In this regard, an initiator entity may determine that outgoing data is to be transmitted to a target entity. The initiator entity may transmit, to the target entity, a solicited push request requesting the outgoing data to be placed at the target entity. In response to the solicited push request, the initiator entity may receive a push grant from the target entity. In response to the push grant, the initiator entity may transmit to the target entity the outgoing data to be placed at the target entity.

METHOD AND APPARATUS FOR MEDIA PROPERTY OR CHARACTERISTIC CONTROL IN A MEDIA SYSTEM (18324409)

Main Inventor

Liang Zhao

Brief explanation

The patent application describes a method for adjusting the multimedia presentation characteristic of a peripheral device based on user input. Here is a simplified explanation of the abstract:

• The control device transmits a portion of content that represents a multimedia presentation characteristic of a peripheral device.
• If the adjustment of the signal is below a certain threshold, the control device receives a user input indicating a request to change the multimedia presentation characteristic.
• In response to the user input, the control device adjusts another portion of the signal and transmits it as part of the content.
• If the adjustment of the signal is at or above the threshold, the control device receives a second instance of the user input.
• In response to the second user input, the control device transmits a communication signal to the peripheral device to adjust its control of the output of the multimedia presentation characteristic.

Potential applications of this technology:

• Home entertainment systems: Users can easily adjust the audio or video settings of their TVs or speakers using a control device.
• Virtual reality systems: Users can change the visual or auditory settings of their VR headsets or accessories for a more immersive experience.
• Conference room setups: Participants can control the audio and video settings of projectors or displays to optimize their presentations.

Problems solved by this technology:

• Simplifies the process of adjusting multimedia presentation characteristics by allowing users to make changes through a control device.
• Eliminates the need for users to manually adjust settings on peripheral devices, making it more convenient and user-friendly.

Benefits of this technology:

• Improved user experience: Users can easily customize the multimedia presentation characteristics to suit their preferences.
• Time-saving: Users can make adjustments without having to physically interact with the peripheral device.
• Greater control: Users have the ability to fine-tune the multimedia presentation characteristics to achieve the desired output.

Abstract

A method including: transmitting, by a control device, a first portion of content comprising a first portion of a signal corresponding to a multimedia presentation characteristic of a peripheral device; receiving, when an adjustment of the signal is below an adjustment threshold, a first instance of an input indicating a request to change the multimedia presentation characteristic; in response to receiving the first instance of the input, adjusting a second portion of the signal and transmitting a second portion of the content comprising the adjusted second portion of the signal; receiving, when the adjustment of the signal is at or above the adjustment threshold, a second instance of the input; and transmitting, in response to receiving the second instance of the input, a communication signal to the peripheral device to adjust a peripheral device control of an output of the multimedia presentation characteristic.

SCANNING PROJECTOR DYNAMIC RESOLUTION (18035284)

Main Inventor

Stuart James Myron Nicholson

Brief explanation

The patent application describes a system for reducing graphical bandwidth in a projection display system by adjusting the modulation frequency used to display different types of content. Here are the key points:

• The system receives content for display and identifies the types of content associated with different portions of the received content.
• For each portion of content, an effective resolution is selected based on the identified content type.
• The system modulates one or more light beams directed to a display area of a projection surface at a specific emitter modulation frequency that corresponds to the selected effective resolution for each portion of content.

Potential applications of this technology:

• Projection display systems in theaters, conference rooms, and other large venues.
• Virtual reality and augmented reality headsets.
• Gaming consoles and computer monitors.

Problems solved by this technology:

• Reduces graphical bandwidth, allowing for more efficient transmission and display of content.
• Improves the overall quality and clarity of displayed content.
• Enables better optimization of resources in projection display systems.

Benefits of this technology:

• Enhanced visual experience for viewers.
• More efficient use of bandwidth and resources.
• Improved performance and flexibility in displaying different types of content.

Abstract

Systems, devices, and techniques are provided for reducing graphical bandwidth in a projection display system by modifying a modulation frequency used to display one or more portions of content based on an identified content type of such portions. Content is received for display, and types of content associated with one or more portions of the received content are identified. For each respective portion of content, an effective resolution is selected for displaying the respective portion based on the identified content type for the respective portion, and one or more light beams directed to a respective display area of a projection surface are modulated at a respective emitter modulation frequency that corresponds to the selected effective resolution for the respective portion.

METHODS, SYSTEMS, AND MEDIA FOR MODIFYING THE PRESENTATION OF VIDEO CONTENT ON A USER DEVICE BASED ON A CONSUMPTION OF THE USER DEVICE (18244486)

Main Inventor

Joshua Burkart

Brief explanation

The patent application describes methods, systems, and media for modifying the presentation of video content on a user device based on the consumption mode of the device. Here are the key points:

• The method involves receiving a request to present a video content item on a user device.
• The audio portion of the video content item is analyzed to identify a segment with a particular type of audio content (e.g., silent, music, speech, theatrical, etc.).
• The consumption mode of the user device is determined.
• A modified version of the video content item is generated by removing the portion that corresponds to the identified segment of audio content.
• The modified video content item is then presented on the user device.

Potential applications of this technology:

• Customized viewing experience: Users can have a tailored video presentation based on their device's consumption mode.
• Noise reduction: The technology can remove unwanted audio segments, such as silence or background noise, to enhance the viewing experience.
• Personalized content: Users can have specific types of audio content filtered out based on their preferences.

Problems solved by this technology:

• Inconsistent viewing experiences: Different devices may have different capabilities or limitations, and this technology ensures that the video content is adapted accordingly.
• Unwanted audio distractions: Users may want to remove certain types of audio content that they find distracting or irrelevant.

Benefits of this technology:

• Improved user experience: Users can enjoy a video presentation that is optimized for their device and preferences.
• Enhanced content consumption: By removing unwanted audio segments, users can focus on the relevant parts of the video content.
• Increased accessibility: Users with specific audio preferences or limitations can have a more accessible video experience.

Abstract

Methods, systems, and media for modifying the presentation of video content on a user device based on a consumption mode of the user device are provided. In some embodiments, a computer-implemented method comprises: receiving, from a user device, a request to present a video content item, wherein the video content item includes an audio portion; identifying a segment of the audio portion as including a particular type of audio content (e.g., silent content, music content, speech content, theatrical content, non-musical content, etc.); determining a consumption mode associated with the user device; generating a modified video content item based on the consumption mode associated with the user device by removing a portion of the video content item that corresponds to the segment of the audio portion identified as including the particular type of audio content; and causing the modified video content item to be presented on the user device.

TIME MARKING CHAPTERS IN MEDIA ITEMS AT A PLATFORM USING MACHINE-LEARNING (18244625)

Main Inventor

Chenjie Gu

Brief explanation

Methods and systems are disclosed for using machine learning to mark the time of media items on a platform. The system takes an identified media item as input and uses a machine learning model to generate one or more outputs. These outputs include time marks that identify each content segment of the media item. Each content segment is associated with a segment start indicator on a timeline of the media item. The system then determines the resulting duration of the combination of content segments for which time marks were obtained. If the resulting duration is less than the duration of the media item, the system provides further inputs to the machine learning model.

• The system uses machine learning to mark the time of content segments in media items.
• It takes an identified media item as input and generates time marks for each content segment using a machine learning model.
• The system determines the resulting duration of the content segments and provides further inputs to the machine learning model if the resulting duration is less than the duration of the media item.

Potential Applications

• This technology can be used in video editing software to automatically mark the time of different scenes or segments in a video.
• It can be applied in content management systems to categorize and organize media items based on their content segments.
• This technology can also be used in recommendation systems to analyze the duration and content segments of media items and provide personalized recommendations to users.

Problems Solved

• Manual marking of content segments in media items can be time-consuming and prone to errors. This technology automates the process using machine learning, saving time and improving accuracy.
• It solves the problem of identifying and categorizing different content segments within a media item, which can be challenging without automated tools.
• This technology addresses the issue of managing and organizing large amounts of media items by providing a systematic way to mark and analyze their content segments.

Benefits

• The use of machine learning reduces the need for manual effort in marking content segments, making the process more efficient.
• Automated time marking of content segments allows for faster editing and organization of media items.
• This technology enables more accurate analysis and categorization of media items based on their content segments, leading to improved recommendation systems and content management.

Abstract

Methods and systems for time marking of media items at a platform using machine-learning are provided herein. An indication of a identified media item is provided as input to a machine-learning model and one or more outputs of the machine-learning model is obtained. The one or more obtained outputs comprise time marks identifying each of the plurality of content segments of the media item. Each of the plurality of content segments is associated with a segment start indicator for a timeline of the media item. A resulting duration is determined of a combination of the plurality of content segments for which the time marks were obtained from the one or more of outputs of the machine-learning model. Responsive to determining that the resulting duration is less than the duration of the media item, one or more further inputs is provided to the machine learning model.

TELEVISION RELATED SEARCHING (18244455)

Main Inventor

Vincent Dureau

Brief explanation

The patent application describes a computer-implemented method for improving search suggestions and search results related to television programming. Here is a simplified explanation of the abstract:

• The method involves identifying metadata (information about the TV programming) displayed on a device.
• Keywords are extracted from the metadata.
• Multiple search suggestions are generated based on these keywords.
• First search results are generated based on the search suggestions.
• The search suggestions and first search results are presented together on the device.

Potential Applications

This technology can have various applications in the field of television programming and search engines. Some potential applications include:

• Enhancing the user experience by providing more relevant search suggestions and results while watching TV.
• Improving the efficiency of searching for specific TV shows, episodes, or related content.
• Assisting in content discovery by suggesting related shows or topics based on the metadata.

Problems Solved

The technology addresses several problems related to searching for TV programming:

• Inadequate search suggestions: Traditional search engines may not provide accurate or relevant suggestions based on TV metadata.
• Limited search results: Users may struggle to find desired TV content due to limited or irrelevant search results.
• Inefficient content discovery: Without effective search suggestions, users may miss out on discovering related TV shows or topics.

Benefits

The use of this technology offers several benefits:

• Enhanced user experience: Users can easily find relevant TV programming based on improved search suggestions and results.
• Time-saving: The method streamlines the search process by presenting both suggestions and initial results together.
• Improved content discovery: Users can explore related TV shows or topics based on the presented search suggestions.
• More efficient searching: The extraction of keywords from metadata helps in generating more accurate search suggestions and results.

Abstract

The subject matter of this specification can be implemented in, among other things, a computer-implemented method that includes identifying metadata related to television programming being presented on a display device. The method further includes extracting one or more keywords from the metadata. The method further includes generating multiple search suggestions based on the keywords and first search results based on one or more of the search suggestions. The method further includes presenting the search suggestions and the first search results together on the display device.

Systems, Apparatus, and Methods for Improving Composition of Images (18251655)

Main Inventor

Yu-Chuan SU

Brief explanation

The patent application describes a method for improving the composition of images using machine learning models and user device movement. Here is a simplified explanation of the abstract:

• The method starts by receiving a first preview image of a scene.
• The first preview image is processed using a machine learning model to determine the best direction for the user device to move based on the composition score of the first preview image and the composition scores of other candidate images.
• The user device movement is detected, and a second preview image of the scene is received after the movement.
• The process can be repeated to continuously improve the composition of the images.

Potential applications of this technology:

• Photography: This technology can be used in cameras or smartphones to help users capture better composed photos by providing real-time feedback and suggestions for device movement.
• Videography: It can also be applied to video recording, assisting users in capturing well-composed videos by guiding their device movement.
• Augmented Reality: This technology can enhance the composition of augmented reality experiences by suggesting optimal device positions for capturing virtual objects in real-world scenes.

Problems solved by this technology:

• Composition improvement: The method addresses the challenge of capturing well-composed images by providing guidance on device movement based on composition scores.
• Real-time feedback: Users can receive immediate feedback on the composition of their images, allowing them to make adjustments and capture better shots.

Benefits of this technology:

• Enhanced user experience: Users can capture visually appealing images with improved composition, leading to more satisfying photography or videography experiences.
• Time-saving: The method eliminates the need for manual trial and error in adjusting device positions, saving time and effort for users.
• Learning-based approach: By utilizing machine learning models, the method can continuously improve its composition suggestions based on user feedback and data.

Abstract

Systems, apparatus, and methods are presented for improving the composition of images. One method includes receiving a first preview image of a scene and processing the first preview image using a first machine learning model to determine at least one direction to move a user device based on a composition score of the first preview image and at least one composition score of a plurality of candidate images. The method also includes detecting movement of the user device and receiving a second preview image of the scene after movement of the user device.

MULTI-FACTOR AUTHENTICATION AND ACCESS CONTROL IN A VEHICULAR ENVIRONMENT (18367868)

Main Inventor

Haris Ramic

Brief explanation

The patent application describes a digital assistant application that can receive sensor signals from a vehicle and determine when someone enters the vehicle. It can then receive authentication input signals from various sensors in the vehicle and determine the authentication states based on these signals and authentication credentials. Based on the authentication states, the application can identify the access permission level and provide access to a subset of functionalities available in the vehicle.

• A digital assistant application for vehicles that can determine when someone enters the vehicle
• Receives authentication input signals from various sensors in the vehicle
• Determines authentication states based on the input signals and authentication credentials
• Identifies the access permission level based on the authentication states
• Provides access to a subset of functionalities available in the vehicle based on the access permission level

Potential Applications

• Vehicle security systems
• Personalized vehicle settings and preferences
• Car sharing services with different access levels for different users

Problems Solved

• Ensures only authorized individuals can access certain functionalities in a vehicle
• Provides a secure and personalized experience for vehicle users
• Helps prevent unauthorized use of vehicles

Benefits

• Improved vehicle security and access control
• Customizable and personalized user experience
• Enhanced safety by limiting access to certain functionalities

Abstract

The systems and methods described herein can include a digital assistant application that receives sensor signals from sensors installed in a vehicle and determines an entry event into the vehicle. The digital assistant application can receive, responsive to the entry event into the vehicle, a plurality authentication input signals from a plurality of sensors associated with the vehicle. The digital assistant application can determine a plurality of authentication states based on the plurality of authentication input signals and a plurality of authentication credentials. The digital assistant application can identify an access permission level of a plurality of access permission levels based at least in part on the plurality of identifies authentication states. The digital assistant application can identify, responsive to the access permission level, a subset of a set of functionalities available via the vehicle, and provide vehicular access to the subset of functionalities.