Video Behavior Prediction Using Object-Relationship Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing behavior recognition technologies only recognize behaviors that have already occurred, making it difficult to take timely countermeasures, and fail to predict future situations where interventions may be necessary.
Innovation Solution
An information processing program and apparatus that analyzes video data to identify relationships between objects and individuals, combines facial expressions and behaviors using machine learning models, and predicts future actions or states through Bayesian inference, enabling proactive detection of potential issues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If behavior recognition technology is used to recognize current or past behaviors, then behavior identification accuracy is improved, but the ability to detect future situations requiring countermeasures deteriorates
Solution Approach 1:
The system performs preliminary action by predicting future behaviors before they actually occur. The behavior prediction model analyzes current and past behaviors to forecast future actions, allowing countermeasures to be taken in advance rather than after the behavior has already happened, thus resolving the time loss problem while maintaining identification accuracy
2Device complexity
If only current and past behavior data is analyzed, then processing simplicity is maintained, but the ability to predict future situations deteriorates
Solution Approach 1:
The system performs preliminary analysis by using the behavior prediction model to forecast future behaviors based on current and past patterns. This preliminary action enables the system to detect future situations requiring countermeasures before they occur, improving reliability without requiring complex real-time intervention systems
Solution Approach 2:
The system implements feedback by continuously monitoring actual behaviors and comparing them with predictions from the behavior prediction model. This feedback loop allows the model to be refined and improved over time, enhancing future situation detection capability while maintaining manageable processing complexity through iterative learning
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An information processing apparatus (10) acquires video data that includes target objects including a person and an object, first identifies a relationship between the target objects in the acquired video data, by using graph data that indicates a relationship between target objects and that is stored in a storage, second identifies a behavior of the person in the video data by using a feature value of the person included in the acquired video data, and predicts one of a future behavior and a future state of the person by inputting the identified behavior of the person and the identified relationship to a machine learning model.