Video-Based Relationship Graph Modeling for Real-Time Interaction Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Evaluating the strength of relationships between individuals, especially in real-time, is difficult, particularly for those with impairments affecting verbal and non-verbal communication abilities, and existing methods lack the granularity to determine interaction-based adjustments.
Innovation Solution
A system and method using computer-implemented machine learning models to analyze digital video data, create relationship graphs, and generate actions to adjust relationships by simulating changes to improve interaction outcomes, enabling real-time analysis and prediction of actions for relationship adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If video data analysis with machine learning models is used to evaluate relationships in real-time, then measurement precision and automation extent are improved, but device complexity and computational requirements increase
Solution Approach 1:
The system segments the relationship evaluation process into distinct components: video data acquisition, feature extraction using first ML model, relationship graph creation, simulation using second ML model, and action generation. This modular segmentation allows each component to be optimized independently while maintaining overall system precision.
Solution Approach 2:
The patent introduces relationship graphs as an intermediary representation that bridges video data and evaluation results. The graphs serve as a structured intermediate format that captures interaction patterns, enabling precise relationship measurement without requiring direct complex analysis of raw video data throughout the entire process.
2Measurement precision
If detailed interaction analysis is performed to improve relationship evaluation granularity, then measurement precision is improved, but loss of time and computational resources increase
Solution Approach 1:
The system performs preliminary action by pre-processing video data to extract relevant features using the first machine learning model before conducting the actual relationship analysis. This preliminary feature extraction step prepares the data in advance, enabling faster and more granular interaction analysis when needed without repeating full processing.
Solution Approach 2:
The patent implements dynamic analysis by adjusting the level of granularity based on the simulation needs. The relationship graphs and simulation process allow the system to focus computational resources on specific interactions that most impact relationship outcomes, rather than uniformly analyzing all interactions at maximum detail.
3Productivity
If simulation and prediction processes are used to generate relationship adjustment actions, then productivity and outcome improvement are enhanced, but device complexity and processing requirements increase
Solution Approach 1:
The system implements feedback through the simulation process, where the second machine learning model predicts how potential actions will affect relationship graphs. This feedback loop allows the system to evaluate multiple action scenarios and select the most effective adjustments, improving productivity through data-driven decision-making rather than trial-and-error approaches.
Solution Approach 2:
The patent uses copying by creating simulated versions of relationship graphs to test potential actions without affecting actual relationships. These copied representations allow the system to explore multiple scenarios and predict outcomes efficiently, reducing the need for complex real-time adjustments while maintaining high productivity.
Data Source
AI summary
A method includes acquiring digital video data that portrays an interacting event, identifying a plurality of features in the digital video data with a first computer-implemented machine learning model, analyzing the plurality of features to create a baseline relationship graph, determining a target relationship graph, generating one or more actions for increasing similarity between the baseline relationship graph and the target relationship graph, and outputting the one or more actions by a user interface. The one or more actions are generated using a simulator, a second computer-implemented machine learning model, and a plurality of actions. The second computer-implemented machine learning model is configured to relate actions of the plurality of actions to changes to relationship graphs, the simulator is configured to simulate changes to the baseline relationship graph using the second computer-implemented machine learning model and the plurality of actions.