NVIDIA Corporation patent applications on June 13th, 2024
Contents
NVIDIA Patent Applications: Cutting-Edge Innovation in AI, Graphics, and Autonomous Systems (June 13, 2024)
Overview of NVIDIA's Latest Patent Filings
On June 13th, 2024, NVIDIA Corporation, a leader in GPU technology and AI computing, submitted 15 new patent applications. These filings showcase NVIDIA's commitment to innovation across various technology sectors, including artificial intelligence, graphics processing, autonomous systems, and 5G technologies.
Key Areas of NVIDIA Patent Applications
AI and Machine Learning
NVIDIA's focus on AI is evident in several patent applications:
- Domain-Customizable Models for Conversational AI Systems
- Involves training large language models (LLMs) for specific domains
- Allows for separate training of model parts associated with different domains
- Enhances AI adaptability and efficiency in various applications
- Neural Vector Fields for 3D Shape Generation
- Introduces a vector field decoder neural network for high-quality 3D shape synthesis
- Capable of zero-shot generation, creating shapes not included in training data
- Potential applications in 3D content creation and CAD modeling
Graphics and Image Processing
NVIDIA continues to push boundaries in graphics technology:
- Luminance-Preserving and Temporally Stable Daltonization
- Addresses color vision deficiency (CVD) issues in image perception
- Maintains consistent color mapping across video frames, reducing flickering
- Improves visual experience for individuals with CVD
- Transformers as Neural Renderers
- Utilizes machine learning to generate views of 3D objects
- Can determine paths of motion based on color values
- Potential applications in computer graphics and robotics
Autonomous Systems and Computer Vision
Several patents focus on improving autonomous vehicle technology:
- Disturbance Compensation for Autonomous Systems
- Enhances control systems for autonomous vehicles
- Uses sensor data to determine system state and generate disturbance data
- Improves responsiveness and stability in autonomous operations
- Object Detection and Confidence for Autonomous Driving
- Develops advanced object detection techniques using machine learning
- Introduces methods for determining ground truth data for training object detectors
- Enhances safety and reliability in autonomous driving systems
5G and Communication Technologies
NVIDIA is also innovating in the realm of 5G technologies:
- Application Programming Interface for 5G-NR Storage Allocation
- Facilitates lock-free data sharing between 5G-NR processes
- Improves efficiency in 5G wireless communication computations
- Enhances performance of 5G network infrastructure
Implications of NVIDIA's Patent Applications
These patent applications demonstrate NVIDIA's commitment to advancing technology across multiple domains. The innovations in AI and machine learning could lead to more efficient and adaptable AI systems, while advancements in graphics processing may result in more realistic and accessible visual experiences.
The focus on autonomous systems and computer vision suggests NVIDIA's continued investment in the future of transportation and robotics. Meanwhile, the company's work on 5G technologies indicates its aim to play a significant role in the evolving landscape of wireless communications.
Conclusion
NVIDIA's latest patent applications reflect the company's broad technological aspirations and its potential to shape the future of computing, AI, and autonomous systems. As these innovations move from patent applications to practical implementations, they could significantly impact various industries and consumer experiences.
Stay tuned for more updates on NVIDIA's patent applications and their potential real-world applications.
NVIDIA Corporation: 15 patent applications
NVIDIA Corporation has applied for patents in the areas of G06T7/90 (4), G06F9/54 (3), H04N23/88 (2), B60W30/18 (1), G06T15/06 (1) G06F9/544 (3), H04N23/88 (2), B60W30/18 (1), G01S7/417 (1), G06F8/65 (1)
With keywords such as: data, based, machine, between, learning, systems, techniques, processes, image, and color in patent application abstracts.
Patent Applications by NVIDIA Corporation
Inventor(s): Mohammed Nasir of San Jose CA (US) for nvidia corporation, Vishal Murali of Santa Clara CA (US) for nvidia corporation, Yue Sun of Danville CA (US) for nvidia corporation
IPC Code(s): B60W30/18, B60W40/10
CPC Code(s): B60W30/18
Abstract: the present disclosure relates to determining an observed state of a system based at least on sensor data generated using one or more sensors of the system. the present disclosure further relates to generating disturbance data based at least on comparing an estimated state of the system with the observed state of the system. the present disclosure further relates to updating one or more disturbance terms of a state space formulation based at least on the disturbance data. the present disclosure further relates to generating, based at least on the state space formulation, a control command that directs one or more operations of the system according to plan data indicative of a plan for completing one or more tasks of the system.
Inventor(s): Tommi Koivisto of Uusimaa (FI) for nvidia corporation, Pekka Janis of Uusimaa (FI) for nvidia corporation, Tero Kuosmanen of Uusimaa (FI) for nvidia corporation, Timo Roman of Uusimaa (FI) for nvidia corporation, Sriya Sarathy of Santa Clara CA (US) for nvidia corporation, William Zhang of Los Altos CA (US) for nvidia corporation, Nizar Assaf of Santa Clara CA (US) for nvidia corporation, Colin Tracey of Santa Clara CA (US) for nvidia corporation
IPC Code(s): G01S7/41, B60W50/00, G01S7/48, G01S13/86, G01S13/931, G01S17/931, G06F16/35, G06F18/21, G06F18/214, G06F18/23, G06F18/2413, G06N3/044, G06N3/045, G06N3/047, G06N3/048, G06N3/084, G06N20/00, G06V10/20, G06V10/44, G06V10/46, G06V10/762, G06V10/764, G06V10/77, G06V10/774, G06V20/58
CPC Code(s): G01S7/417
Abstract: in various examples, detected object data representative of locations of detected objects in a field of view may be determined. one or more clusters of the detected objects may be generated based at least in part on the locations and features of the cluster may be determined for use as inputs to a machine learning model(s). a confidence score, computed by the machine learning model(s) based at least in part on the inputs, may be received, where the confidence score may be representative of a probability that the cluster corresponds to an object depicted at least partially in the field of view. further examples provide approaches for determining ground truth data for training object detectors, such as for determining coverage values for ground truth objects using associated shapes, and for determining soft coverage values for ground truth objects.
Inventor(s): Vishvesh VIJAYWARGIYA of Bangalore (IN) for nvidia corporation, Lalit ADITHYA V of Bangalore (IN) for nvidia corporation, Krishnan DURAISAMY of San Jose CA (US) for nvidia corporation, Rohit RAJANI of Shajapur (IN) for nvidia corporation, Gopi VADLAMUDI of Fremont CA (US) for nvidia corporation, Andrew STOCK of Placentia CA (US) for nvidia corporation, Alexander PELAVIN of Palo Alto CA (US) for nvidia corporation, Shivam MISHRA of Sacramento CA (US) for nvidia corporation, Prathik KOTIAN of Coimbatore (IN) for nvidia corporation
IPC Code(s): G06F8/65
CPC Code(s): G06F8/65
Abstract: an application management platform comprising at least a packaging and bundling component, a deployment management component, and an update component. the packaging and bundling component versions, packages, and bundles a plurality of infrastructure components for a remote data center. the deployment management component provisions one or more nodes of the remote data center with the plurality of infrastructure components for an application. the update component monitors available updates to one or more of the plurality of infrastructure components used by the remote data center and facilitates update of the one or more of the plurality of infrastructure components at the remote data center.
Inventor(s): Dominik Wachowicz of Gdansk (PL) for nvidia corporation
IPC Code(s): G06F9/54
CPC Code(s): G06F9/544
Abstract: disclosed are apparatuses, systems, and techniques for efficient parallel execution of multiple processes in real-time streaming and latency-sensitive applications. the techniques include but are not limited to executing in parallel multiple processing threads, storing data output by the multiple processing threads in respective accumulation buffers, and applying an aggregation function to the stored data to generate an aggregated data.
Inventor(s): Jinyou Wu of Shanghai (CN) for nvidia corporation
IPC Code(s): G06F9/54
CPC Code(s): G06F9/544
Abstract: apparatuses, systems, and techniques to perform and facilitate lock-free data sharing between processes performing computations in fifth generation (5g) new radio (nr) wireless communication. in at least one embodiment, one or more statically-sized regions of linked storage locations are pre-allocated, in response to an application programming interface (api), to store 5g-nr information to be shared between one or more processes.
Inventor(s): Jinyou Wu of Shanghai (CN) for nvidia corporation
IPC Code(s): G06F9/54
CPC Code(s): G06F9/544
Abstract: apparatuses, systems, and techniques to perform and facilitate lock-free data sharing between processes performing computations in fifth generation (5g) new radio (nr) wireless communication. in at least one embodiment, 5g-nr information stored by one or more statically-sized regions of linked storage locations is to be invalidated in response to an application programming interface (api).
Inventor(s): Yi Dong of Lexington MA (US) for nvidia corporation, Xianchao Wu of Tokyo (JP) for nvidia corporation
IPC Code(s): G06N5/043, G06F40/40
CPC Code(s): G06N5/043
Abstract: in various examples, systems and methods are disclosed that train a machine learning model(s)—such as a large language model (llm)—for one or more specific domains. in some embodiments, the machine learning model(s) may include at least a base model(s) as well as additional parts, such as additional layers, associated with the domains for which the machine learning model(s) is being trained. as such, the parts of the machine learning model(s) may be trained separately, such that training data associated with a domain is used to train a part of the machine learning model(s) that is associated with the domain without training the other part(s) of the machine learning model(s). the systems and methods may then use these parts when deploying the machine learning model(s), such as by activating and/or deactivating parts based on the input data being processed.
Inventor(s): Johan Pontus Ebelin of Skåne (SE) for nvidia corporation, Cyril Crassin of Courbevoie (FR) for nvidia corporation, Tomas Guy Akenine-Möller of Lund (SE) for nvidia corporation
IPC Code(s): G06T11/00, G06T7/90
CPC Code(s): G06T11/001
Abstract: it is difficult for people with color vision deficiency (cvd) to distinguish between certain colors, e.g., reds and greens may be indistinguishable, causing a loss of information. image recoloring, daltonization, techniques aim to improve the experience for people with cvd. preserving luminance between the original image as seen by a person with normal color vision and someone with a cvd assists in preserving image appearance. conventional algorithms attempt to daltonize images by exploiting the content of the image itself. while this is a suitable idea for an image in isolation, temporal inconsistencies (e.g., flickering) occur when applied to a stream of images, as a color c could be mapped to a color a in one frame and b in another. in contrast, the luminance-preserving technique operates on pixels and provides a consistent mapping and therefore is temporally stable.
20240193848.TRANSFORMERS AS NEURAL RENDERERS_simplified_abstract_(nvidia corporation)
Inventor(s): Yue Wang of Mountain View CA (US) for nvidia corporation, Marco Pavone of Stanford CA (US) for nvidia corporation
IPC Code(s): G06T15/06, G06T7/90
CPC Code(s): G06T15/06
Abstract: apparatuses, systems, and techniques to use one or more machine learning processes to obtain a set of feature values based at least in part on a set of locations along a ray that intersects an object. a color value is obtained based at least in part on the set of feature values. a view of the object may be generated using the color value. a path of motion may be determined based at least in part on the color value and used to cause a device to move.
20240193887.NEURAL VECTOR FIELDS FOR 3D SHAPE GENERATION_simplified_abstract_(nvidia corporation)
Inventor(s): Zekun Hao of New York NY (US) for nvidia corporation, Ming-Yu Liu of San Jose CA (US) for nvidia corporation, Arun Mohanray Mallya of San Jose CA (US) for nvidia corporation
IPC Code(s): G06T19/20
CPC Code(s): G06T19/20
Abstract: synthesis of high-quality 3d shapes with smooth surfaces has various creative and practical use cases, such as 3d content creation and cad modeling. a vector field decoder neural network is trained to predict a generative vector field (gvf) representation of a 3d shape from a latent representation (latent code or feature volume) of the 3d shape. the gvf representation is agnostic to surface orientation, all dimensions of the vector field vary smoothly, the gvf can represent both watertight and non-watertight 3d shapes, and there is a one-to-one mapping between a predicted 3d shape and the ground truth 3d shape (i.e., the mapping is bijective). the vector field decoder can synthesize 3d shapes in multiple categories and can also synthesize 3d shapes for objects that were not included in the training dataset. in other words, the vector field decoder is also capable of zero-shot generation.
Inventor(s): Seema Kumar of Santa Clara CA (US) for nvidia corporation, Ish Chadha of San Jose CA (US) for nvidia corporation
IPC Code(s): H04L7/00
CPC Code(s): H04L7/0083
Abstract: a device includes receiver circuitry to receive incoming signals on a clock lane and data lanes and detection circuitry. the detection circuitry is to monitor the incoming signals on the clock lane, and determine that an incoming pattern of the incoming signals on the clock lane does not correspond to a clock pattern associated with communicating data on the data lanes. the detection circuitry is to initiate a power-down sequence in response to determining that the incoming pattern does not correspond to the clock pattern.
Inventor(s): Jinyou Wu of Shanghai (CN) for nvidia corporation
IPC Code(s): H04L67/1097, H04L67/1001
CPC Code(s): H04L67/1097
Abstract: apparatuses, systems, and techniques to perform and facilitate lock-free data sharing between processes performing computations in fifth generation (5g) new radio (nr) wireless communication. in at least one embodiment, 5g-nr information to be shared between one or more processes is stored by one or more statically-sized regions of linked storage locations in response to an application programming interface (api).
Inventor(s): Douglas Taylor of Santa Clara CA (US) for nvidia corporation, Animesh Khemka of Fremont CA (US) for nvidia corporation
IPC Code(s): H04N23/88, G06T7/90, H04N9/77, H04N23/71, H04N23/76
CPC Code(s): H04N23/88
Abstract: apparatuses, systems, and techniques for white balancing an image are presented. in at least one embodiment, a chromaticity-based weighting function is determined based at least on an estimated scene brightness of the image and applied to exclude or minimize the impact of large colored portions or objects within an image when estimating an illuminant color.
20240196105.FALLBACK MECHANISM FOR AUTO WHITE BALANCING_simplified_abstract_(nvidia corporation)
Inventor(s): Douglas Taylor of Santa Clara CA (US) for nvidia corporation, Animesh Khemka of Fremont CA (US) for nvidia corporation, Eric Dujardin of San Jose CA (US) for nvidia corporation
IPC Code(s): H04N23/88, G01J1/10, G06T7/90
CPC Code(s): H04N23/88
Abstract: improved fallback mechanisms for auto white balancing are presented. in at least one embodiment, white balance correction factors produced by a first white balance technique are blended with white balance correction factors produced by a second white balance technique based on a confidence level in the white balance correction factors produced by the first white balance technique.
Inventor(s): Jinyou Wu of Shanghai (CN) for nvidia corporation
IPC Code(s): H04W8/00, H04W76/30
CPC Code(s): H04W8/00
Abstract: apparatuses, systems, and techniques to perform and facilitate lock-free data sharing between processes performing computations in fifth generation (5g) new radio (nr) wireless communication. in at least one embodiment, one or more statically-sized regions of linked storage locations are deallocated, in response to an application programming interface (api), to free memory used to store 5g-nr information to be shared between one or more processes.
- NVIDIA Corporation
- Patent Applications
- Artificial Intelligence
- Graphics Processing
- Autonomous Systems
- 5G Technology
- B60W30/18
- B60W40/10
- CPC B60W30/18
- Nvidia corporation
- G01S7/41
- B60W50/00
- G01S7/48
- G01S13/86
- G01S13/931
- G01S17/931
- G06F16/35
- G06F18/21
- G06F18/214
- G06F18/23
- G06F18/2413
- G06N3/044
- G06N3/045
- G06N3/047
- G06N3/048
- G06N3/084
- G06N20/00
- G06V10/20
- G06V10/44
- G06V10/46
- G06V10/762
- G06V10/764
- G06V10/77
- G06V10/774
- G06V20/58
- CPC G01S7/417
- G06F8/65
- CPC G06F8/65
- G06F9/54
- CPC G06F9/544
- G06N5/043
- G06F40/40
- CPC G06N5/043
- G06T11/00
- G06T7/90
- CPC G06T11/001
- G06T15/06
- CPC G06T15/06
- G06T19/20
- CPC G06T19/20
- H04L7/00
- CPC H04L7/0083
- H04L67/1097
- H04L67/1001
- CPC H04L67/1097
- H04N23/88
- H04N9/77
- H04N23/71
- H04N23/76
- CPC H04N23/88
- G01J1/10
- H04W8/00
- H04W76/30
- CPC H04W8/00
