Qualcomm incorporated (20240233331). MESSAGE PASSING NETWORK BASED OBJECT SIGNATURE FOR OBJECT TRACKING simplified abstract
Contents
MESSAGE PASSING NETWORK BASED OBJECT SIGNATURE FOR OBJECT TRACKING
Organization Name
Inventor(s)
Avdhut Joshi of San Marcos CA (US)
Sajal Maheshwari of Seattle WA (US)
Ahmed Kamel Sadek of San Diego CA (US)
MESSAGE PASSING NETWORK BASED OBJECT SIGNATURE FOR OBJECT TRACKING - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240233331 titled 'MESSAGE PASSING NETWORK BASED OBJECT SIGNATURE FOR OBJECT TRACKING
Simplified Explanation: The patent application describes a system for processing image data, which involves computing initial embeddings from multiple images, constructing a graph based on these embeddings, performing message passing steps on the graph to generate final embeddings, classifying objects in the images using a classification engine, and computing a classification loss based on the classification results.
- The system computes initial embeddings from a set of images.
- It constructs a graph with nodes representing the initial embeddings.
- The system performs message passing steps on the graph to generate final embeddings.
- Objects in the images are classified using a classification engine based on the final embeddings.
- A classification loss is computed based on the classification results.
Potential Applications: 1. Image recognition and classification systems. 2. Object detection in images. 3. Visual search engines. 4. Automated image tagging and labeling.
Problems Solved: 1. Efficient processing of image data. 2. Accurate object classification in images. 3. Streamlining image analysis tasks. 4. Enhancing image-based search capabilities.
Benefits: 1. Improved accuracy in object classification. 2. Faster processing of large image datasets. 3. Enhanced image search functionality. 4. Automation of image analysis tasks.
Commercial Applications: The technology can be utilized in various industries such as e-commerce for product recognition, security for surveillance systems, healthcare for medical image analysis, and social media for content moderation.
Questions about Image Data Processing: 1. How does the system compute initial embeddings from images? 2. What is the significance of constructing a graph based on the initial embeddings?
Frequently Updated Research: Researchers are continually exploring advancements in image processing algorithms to enhance object recognition and classification accuracy. Stay updated on recent developments in this field to leverage the latest technologies for image data processing.
Original Abstract Submitted
disclosed are systems, apparatuses, processes, and computer-readable media for processing image data. for example, an apparatus can compute initial embeddings from a plurality of images. the apparatus can construct a graph comprising nodes representing the initial embeddings. the apparatus can further perform, based on the graph, a plurality of message passing steps successively to generate final embeddings. the apparatus can classify, using a classification engine, one or more objects in each of the plurality of images based on the final embeddings. the apparatus can further compute a classification loss based on the classifying of the one or more objects.
