Visual Recognition Using Social Network Co-occurrences
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current visual recognition technologies primarily focus on image content and ignore additional data that could be leveraged for improved performance, such as social relationships and semantic information.
Innovation Solution
A multi-label modeling approach using a multiple kernel learning framework that jointly models content, semantic, and social network information, incorporating co-occurrences of individuals in images to infer relationships and annotations, and trains a joint kernel to identify annotations and recommendations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If visual recognition focuses only on image content, then the system is simple and fast, but recognition accuracy and annotation quality deteriorate due to ignoring useful additional data
Solution Approach 1:
The patent combines multiple data sources including image content, social network information, and semantic relationships into a unified visual recognition system. The system merges these different types of data to jointly determine annotations, improving recognition accuracy by leveraging complementary information from multiple sources rather than relying on image content alone.
Solution Approach 2:
The visual recognition system is designed to handle multiple functions simultaneously: it processes image content analysis, social network relationship extraction, semantic information integration, and annotation generation. This multi-functional approach allows the system to utilize various data types for comprehensive image understanding and annotation.
2Manufacturing precision
If the system incorporates multiple data sources including social networks and semantics, then annotation quality improves, but computational complexity and processing time increase
Solution Approach 1:
The system performs preliminary processing of social network information and semantic relationships during an offline training phase, pre-computing features and relationships that will be used during annotation. This preliminary action reduces the computational burden during actual annotation processing, thereby reducing processing time while maintaining high annotation quality.
Solution Approach 2:
The annotation process is divided into separate modules: image content analysis, social network relationship extraction, semantic information processing, and integrated annotation generation. This segmentation allows each module to be optimized independently and processed in parallel where possible, reducing overall processing time while maintaining comprehensive annotation quality.
Data Source
AI summary
System, method and architecture for providing improved visual recognition by modeling visual content, semantic content and an implicit social network representing individuals depicted in a collection of content, such as visual images, photographs, etc., which network may be determined based on co-occurrences of individuals represented by the content, and/or other data linking the individuals. In accordance with one or more embodiments, using images as an example, a relationship structure may comprise an implicit structure, or network, determined from co-occurrences of individuals in the images. A kernel jointly modeling content, semantic and social network information may be built and used in automatic image annotation and/or determination of relationships between individuals, for example.