Emotion Estimation Using Multi-Source Data Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing emotion estimation methods, such as those based on facial expressions, struggle to accurately evaluate various emotions like happiness and surprise, as they often rely on a single data type and fail to provide reliable intensity measurements.
Innovation Solution
An emotion estimating method that acquires both physiology and non-physiology data, using a camera to capture heart rate and facial expressions, calculates arousal and valence levels, and outputs these as points in a two-dimensional model to intuitively represent emotions, with adjustable reliability indicators and reference values based on physical characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If only facial expression data is used for emotion estimation, then the device complexity is reduced, but the measurement precision and reliability of emotion estimation deteriorates
Solution Approach 1:
The patent combines multiple data sources including facial expressions, voice characteristics, and physiological data into a unified emotion estimation system. This merging of diverse data types enables more accurate and reliable emotion detection while maintaining manageable system complexity through integrated processing.
Solution Approach 2:
The system uses a composite approach by integrating multiple types of data (facial, vocal, physiological) to create a comprehensive emotion estimation model. This composite data strategy enhances measurement precision by leveraging the complementary strengths of different data sources.
2Reliability
If multiple types of data are acquired from the subject, then the reliability of emotion estimation is improved, but the device complexity increases
Solution Approach 1:
The emotion estimation apparatus is designed with multi-functionality to handle various data types (facial expressions, voice, physiological signals) through a unified processing framework. This universal approach allows the system to acquire and process multiple data sources without proportionally increasing complexity, as the core processing architecture remains consistent across different input types.
3Measurement precision
If physiology data and non-physiology data are combined, then the accuracy of emotion intensity measurement is improved, but the difficulty of detecting and measuring increases
Solution Approach 1:
The patent segments the emotion estimation process into distinct components: facial expression analysis, voice characteristic analysis, and physiological data processing. Each segment handles specific data types independently before integrating results, which reduces the overall difficulty by breaking down the complex task of combining multiple data sources into manageable parts.
Solution Approach 2:
The system measures emotions in a two-dimensional space defined by arousal and valence, transforming complex multi-source data into a standardized coordinate system. This dimensional transformation simplifies the integration of physiology and non-physiology data by mapping them to common emotional dimensions, making the combined measurement more tractable.
Data Source
AI summary
An emotion estimating method includes: acquiring first data and second data from the subject, the first data corresponding to physiology data, and the second data corresponding to one of physiology data different from the first data and non-physiology data; calculating a first value and a second value based on the acquired first data and second data, the first value indicating a degree of arousal of the subject and the second value indicating a degree of valence of the subject; estimating the emotion of the subject, by using the calculated first value and second value and based on a predetermined association of people's emotions with a degree of arousal and a degree of valence of the people, the predetermined association being pre-stored in a memory; and outputting information indicating the estimated emotion.