Eye Feature Inference via Bayesian Deep Belief Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current computer systems face challenges in accurately inferring human mental states from behavioral data, as facial expressions and other voluntary actions can be deceptive, and existing machine learning methods have not fully leveraged the most informative and universally expressive features of human eye behavior.
Innovation Solution
A computational model using Bayesian deep belief networks and computer vision to extract and analyze eye features from video data, such as pupil dilation, blink rate, and gaze behavior, to make real-time inferences about cognitive and emotional states, providing a more robust and accurate method for mental state inference.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If facial expressions and voluntary actions are used to infer mental states, then the system can capture rich behavioral data, but the accuracy of mental state inference deteriorates because these behaviors can be deceptive
Solution Approach 1:
The patent extracts and isolates eye behavior features (pupil dilation, blink rate, gaze patterns) from the overall behavioral data set. By focusing specifically on involuntary eye movements rather than voluntary facial expressions, the system extracts the reliable signal while leaving out the deceptive elements of human behavior.
Solution Approach 2:
The patent applies local quality by treating different behavioral features with different levels of trust. Eye behaviors are assigned higher reliability weights compared to facial expressions, creating a differentiated quality assessment where involuntary physiological signals carry more evidentiary value in mental state inference.
2Productivity
If existing machine learning methods are used to analyze behavioral data, then the system can process large amounts of data, but the accuracy of mental state discrimination deteriorates because existing methods have not fully leveraged eye behavior features
Solution Approach 1:
The patent transforms raw video data into specific quantitative eye behavior parameters (pupil diameter in pixels, blink rate per minute, gaze fixation duration). By changing the parameter representation from general behavioral descriptors to precise ocular metrics, the system enables more accurate mental state discrimination while maintaining efficient data processing.
Solution Approach 2:
The patent segments the analysis into distinct eye behavior components (pupil response, blinking patterns, saccadic movements, fixation duration) rather than treating behavioral data as a unified signal. This segmentation allows the machine learning model to process each feature type with appropriate algorithms, improving overall discrimination accuracy.
3Reliability
If comprehensive behavioral features are collected to improve mental state inference accuracy, then the information quality improves, but the system complexity increases
Solution Approach 1:
The patent introduces an intermediary layer of computer vision algorithms that automatically extract eye behavior features from standard video feeds. This intermediary processing stage transforms complex raw video data into structured eye movement parameters, achieving high inference accuracy without requiring complex specialized hardware.
Solution Approach 2:
The patent uses standard video cameras to capture eye behavior rather than specialized eye-tracking hardware. By copying the functionality of expensive specialized devices using readily available components, the system achieves comprehensive behavioral feature collection while minimizing device complexity and cost.
Data Source
AI summary
A method of discovering relationships between eye movements and cognitive and/or emotional responses of a user starts by engaging the user in a task having visual stimuli via an electronic display configured to elicit a predicted specific cognitive and/or emotional response from the user. The visual stimuli are varied to elicit the predicted specific cognitive and/or emotional response from the user. A camera films an eye of the user. A first time series of eye movements is recorded by the camera. A computing device compares the eye movements from the first time series and the tasks and identifies at least one relationship between eye movements that correlate to the actual specific cognitive and/or emotional response.