Video Character Context Extraction via Temporal Graph Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies struggle to effectively extract and utilize the contextual importance of a person appearing in a video, including their role and interactions with other characters.
Innovation Solution
A method and device that obtain appearance time information for characters in a video, extract context related to the characters, generate graphs representing the context, and classify characters based on these graphs to identify their contextual importance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If existing image processing technologies are used to extract information from video, then basic character recognition and movement detection can be achieved, but the contextual importance and role of characters cannot be effectively extracted
Solution Approach 1:
The video is segmented into multiple time points based on character appearance, and at each time point, a graph is constructed to represent relationships between characters and objects. This segmentation allows the system to process and analyze contextual information at discrete temporal intervals, making the extraction of contextual importance manageable and systematic.
Solution Approach 2:
The patent introduces a temporal dimension to the analysis by constructing graphs at multiple time points corresponding to character appearances. This transforms the static image analysis into a dynamic temporal graph analysis, enabling the extraction of contextual information that evolves over time and cannot be captured by single-frame analysis alone.
2Measurement precision
If comprehensive context extraction including interactions and visual-auditory features is performed, then character classification accuracy improves, but processing time increases
Solution Approach 1:
The system pre-extracts visual and auditory features from video frames before constructing the graphs. By preparing these features in advance at multiple time points, the subsequent graph construction and character classification processes can proceed more efficiently, reducing overall processing time while maintaining comprehensive analysis.
Solution Approach 2:
The patent continuously extracts and updates contextual information at multiple time points corresponding to character appearances. This continuous extraction process ensures that all relevant contextual data is captured without gaps, maintaining high classification accuracy while optimizing the extraction process to minimize time loss.
Data Source
AI summary
A method and device for obtaining context associated with a character in a video are disclosed. A method for obtaining a context associated with a character in a video, performed by a device according to one embodiment of the present disclosure, may include obtaining appearance time information for at least one character appearing in a specific video; obtaining a context related to at least one character from a video portion corresponding to a time at which the at least one character appeared based on the appearance time information; generating at least one graph representing a context of the at least one character based on the context related to the at least one character; and classifying the at least one character using the at least one graph.