18152627. HUMAN OBJECT INTERACTION DETECTION USING COMPOSITIONAL MODEL simplified abstract (Accenture Global Solutions Limited)
Contents
HUMAN OBJECT INTERACTION DETECTION USING COMPOSITIONAL MODEL
Organization Name
Accenture Global Solutions Limited
Inventor(s)
David Nguyen of Newark CA (US)
HUMAN OBJECT INTERACTION DETECTION USING COMPOSITIONAL MODEL - A simplified explanation of the abstract
This abstract first appeared for US patent application 18152627 titled 'HUMAN OBJECT INTERACTION DETECTION USING COMPOSITIONAL MODEL
The patent application describes a method for analyzing images to determine human-object interactions (HOIs) using visual and semantic embeddings.
- Receiving an image and extracting a visual HOI and visual embeddings.
- Obtaining semantic HOIs and semantic embeddings based on the subject, object, and verbs.
- Processing visual and semantic embeddings to generate transition embeddings.
- Calculating scores based on transition embeddings to determine similarity between visual and semantic HOIs.
- Predicting HOIs within the image based on the scores.
Potential Applications: This technology can be used in image recognition systems, video surveillance, and augmented reality applications.
Problems Solved: This technology addresses the challenge of accurately identifying human-object interactions in images.
Benefits: Improved accuracy in recognizing and understanding human-object interactions in visual data. Enhanced performance of image analysis systems. Increased efficiency in processing and interpreting visual information.
Commercial Applications: Title: Enhanced Image Recognition Technology for Improved Visual Analysis This technology can be utilized in security systems, retail analytics, and industrial automation for improved object detection and tracking.
Prior Art: Researchers can explore prior art related to image recognition, visual embeddings, and semantic analysis techniques in the field of computer vision.
Frequently Updated Research: Stay updated on advancements in deep learning models for image analysis, semantic embedding techniques, and applications of HOI detection in computer vision research.
Questions about Image Recognition Technology: 1. How does this technology improve the accuracy of identifying human-object interactions in images? 2. What are the potential commercial applications of this enhanced image recognition technology?
Original Abstract Submitted
Implementations include actions of receiving an image; extracting a visual HOI and a set of visual embeddings, the visual HOI indicating a subject and an object; obtaining, using a vector library, a set of semantic HOIs and sets of semantic embeddings based on the subject, the object and a set of verbs included in the vector library, each set of semantic embeddings corresponding to a semantic HOI; processing, by a compositional model, the set of visual embeddings to provide a set of transition visual embeddings; processing the sets of semantic embeddings to provide respective sets of transition semantic embeddings; determining a set of scores based on the set of transition visual embeddings and the sets of transition semantic embeddings, each score representing a degree of similarity between the visual HOI and a semantic HOI; and determining at least one predicted HOI represented within the image based on the scores.
