Facial Expression Analysis System Using Affect Space Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for measuring human responses to visual stimuli, such as facial expressions, are limited by requiring ideal imaging conditions and struggle to accurately capture natural behavior in public or retail settings with varying lighting and face orientations.
Innovation Solution
A system utilizing emotion-sensitive feature extraction and learning machines to analyze facial muscle actions from gray-level or color images, employing a set of gradient filters to detect weak contours and align facial features, mapping these actions to affective states and ultimately predicting responses like purchase decisions or opinions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional facial image analysis methods are used, then measurement precision is improved under ideal imaging conditions, but the system fails to accurately capture natural behavior under realistic conditions with varying lighting and face orientations
Solution Approach 1:
The system transforms facial images into affect space parameters (arousal, valence, stance) that are invariant to imaging conditions. By mapping facial expressions to these psychological parameters rather than raw pixel values, the system achieves consistent measurement precision across varying lighting and orientation conditions.
Solution Approach 2:
The patent introduces affect space as an intermediary representation between raw facial images and emotional response measurements. This intermediary layer (with dimensions of arousal, valence, and stance) acts as a bridge that decouples the measurement from specific imaging conditions while preserving the essential emotional information.
2Adaptability or versatility
If the system processes natural behavior in public places, then adaptability is improved, but measurement precision deteriorates due to uncontrolled imaging conditions
Solution Approach 1:
The system changes the parameter space from raw image coordinates to affect space (arousal, valence, stance), making the measurements invariant to the uncontrolled conditions of public settings. This allows accurate emotional response measurement even when lighting, distance, and angle vary.
Solution Approach 2:
The affect space representation serves multiple functions simultaneously: it captures emotional state, compensates for imaging variations, and provides a unified framework for analyzing diverse facial expressions across different contexts and conditions.
3Measurement precision
If detailed facial feature analysis is performed, then measurement precision is improved, but device complexity increases due to the need for multiple processing steps
Solution Approach 1:
The system extracts only the essential emotional information from facial images by mapping to affect space parameters (arousal, valence, stance). This extraction approach discards irrelevant details about specific facial features while retaining the core emotional signal, reducing processing complexity.
Solution Approach 2:
By transforming the problem from detecting specific facial features to measuring affect space parameters, the system simplifies the processing pipeline. The mapping to three-dimensional affect space provides a compact representation that captures emotional content without requiring complex feature-by-feature analysis.
Data Source
AI summary
The present invention is a method and system for measuring human emotional response to visual stimulus, based on the person's facial expressions. Given a detected and tracked human face, it is accurately localized so that the facial features are correctly identified and localized. Face and facial features are localized using the geometrically specialized learning machines. Then the emotion-sensitive features, such as the shapes of the facial features or facial wrinkles, are extracted. The facial muscle actions are estimated using a learning machine trained on the emotion-sensitive features. The instantaneous facial muscle actions are projected to a point in affect space, using the relation between the facial muscle actions and the affective state (arousal, valence, and stance). The series of estimated emotional changes renders a trajectory in affect space, which is further analyzed in relation to the temporal changes in visual stimulus, to determine the response.