Autoencoder Gaze Modeling for Internal State Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to accurately model and estimate internal states, such as concentration levels or emotions, from biological information due to difficulties in reasonably modeling causal relations between biological data and internal states, particularly in driving scenarios and medical diagnostics, leading to unreliable model structures and explanations.
Innovation Solution
An electronic device employing an autoencoder architecture with an encoder and decoder, utilizing machine learning to estimate internal states by reconstructing biological information from environmental and internal state data, adjusting parameters based on reproducibility and probability distribution divergence, to objectively estimate concentration levels and other internal states.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional methods are used to estimate internal states from biological information, then the estimation process can be performed, but the accuracy and reliability of the estimation are insufficient due to unreasonable model structures
Solution Approach 1:
The patent introduces a feedback mechanism where the estimated internal state is used to adjust the model parameters. The system continuously refines the estimation by comparing predicted biological information with actual observed data, thereby improving both accuracy and reliability of the model structure over time
Solution Approach 2:
The patent dynamically adjusts model parameters based on the estimated internal state and observed biological information. By changing parameters adaptively rather than using fixed structures, the system achieves more accurate and reliable internal state estimation across different conditions
2Measurement precision
If complex model structures are used to capture causal relations between biological information and internal states, then estimation capability is enhanced, but the difficulty of modeling and explaining the model increases
Solution Approach 1:
The patent segments the complex modeling task into distinct components: biological information processing, internal state estimation, and causal relation modeling. Each component is handled separately with dedicated processing, reducing overall model complexity while maintaining estimation accuracy
Solution Approach 2:
The patent introduces an intermediary representation layer that bridges biological information and internal states. This intermediate layer simplifies the direct mapping complexity by transforming raw biological data into meaningful features before estimating internal states, making the model more explainable
3Measurement precision
If more biological information and environmental data are collected to improve internal state estimation, then estimation accuracy can be enhanced, but the complexity of data processing and model training increases
Solution Approach 1:
The patent extracts only the most relevant features from the collected biological information and environmental data. By selecting and extracting key features rather than processing all raw data, the system maintains high estimation accuracy while reducing data processing complexity
Solution Approach 2:
The patent performs preliminary processing and filtering of biological information and environmental data before feeding them into the main estimation model. This preliminary action prepares the data in advance, reducing the computational burden during actual estimation and training phases
Data Source
AI summary
An electronic device includes an encoder and a decoder. The encoder is configured to estimate an unknown value on the basis of first biological information including a line of sight of a subject extracted from an image of the subject, subject's environmental information representing an environment of the subject, and subject's internal state information representing an internal state of the subject. The decoder is configured to estimate second biological information including the line of sight of the subject on the basis of the unknown value, the subject's environmental information, and the subject's internal state information. The electronic device adjusts parameters of the encoder and the decoder on the basis of reproducibility of the first biological information from the second biological information.