Use Cases
Use Case 1: Handsfree operation of autofocal glasses by LFI-sensor-based eye-tracking
Motivation and Objectives of the Use Case 1
- Improve the handling, safety and comfort of autofocal glasses while operating during daily activities, e.g. driving in a car or simply walking downstairs.
- Eye-tracking system shall monitor eye-movements and detect reading behavior based on known movement sequences and trigger certain system controls based on detection of these sequences.
- LFI sensor-based eye-tracking systems are the only technology to fit into all-day wearable and aesthetically appealing smartglasses.
- Use Case implementation will showcase the hands-free operation of autofocal glasses by eye-movements.
- People wearing the glasses will be able to read small print intuitively and navigate their surroundings with appropriate visual acuity.
Approach and Demonstrator:
- The demonstrator will be based on a common glasses frame with at least 2 integrated LFI sensor-based eye-tracking systems on both eyes. To cater to anatomical variations in the human populations 3 models with different dimensions are foreseen.
- The integrated eye-tracking system will monitor eye movements and continuously compare the data with known sequences of eye movements associated with reading. A decision tree for activating and deactivating the tunable lenses will be implemented, tested, and optimized.
- The glasses will be tested by various users to validate its more intuitive improvements in comparison to the current approach (VOG/EOG) results.
Aimed Results:
- Autofocal glasses demonstrator with integrated LFI sensor-based eye-tracking system.
- Hands-free operation of the glasses, with the ability to read fine print when desired.
- Reliable autofocal switching (>95% cases) based on viewing angle determined by eye- tracking.
- Robust against downwards gaze when absent of the intent to read.
- Gaze angle detection frequency > 100Hz.
Use Case 2: Cognitive Load & Attention tracking
Motivation and Objectives of the Use Case 2
- Development of an algorithm for cognitive load detection based on multiple features, e.g., pupillary information, microsaccades, visual scan-path analysis etc.
- Integration of the implemented algorithm into an adaptive interface.
- Evaluation against video-based eye-tracking.
Approach and Demonstrator:
- Assessment of cognitive load by consideration of several data streams derived from the novel wearable like pupillary information and eye movement features.
- Utilize fixation, saccade, and microsaccade related features in the model:
- Number of fixations per second
- Saccade characteristics
- Proposed technology's high sampling frequency enables detection of microsaccades:
- Microsaccades are small involuntary eye movements during fixation
- More visually demanding tasks increase microsaccade frequency
- Develop a method for classifying cognitive load with:
- Robust estimation
- High classification accuracy
- Generalizability across participants
- Potential for real-time application
Aimed Results:
- Practically viable model for cognitive load detection allows assessment of the user’s cognitive load across tasks and subjects.
- Method capable of running in real-time based on a sliding window approach.
- Model which allows adaptation of interfaces.
- Accurate and robust cognitive load detection with over 90% of accuracy when three-levels of cognitive load are considered, surpassing the state-of-the-art even in the high-frequency eye-tracking setup.
- Online and generic methods for cognitive load detection that will improve user experience, beyond the proposed eye-tracking system that will be evaluated with other eye-tracking datasets.
- Harmonious synchronization of cognitive load detection with adaptive user interfaces.
- Seamless synchronization of pupillary information with microsaccadic eye characteristics.
Use Case 3:
Motivation and Objectives of the Use Case 3
- Ensure privacy in eye-tracking systems by avoiding the use of camera-based solutions that capture real images of the eyes, which pose legal and ethical challenges.
- Demonstrate that VIVA’s eye-tracking technology, which is not image-based, can mitigate privacy concerns while maintaining high performance in person identification tasks.
- Investigate whether accurate person identification can be achieved using non-intrusive data streams instead of conventional scene or eye cameras.
Approach and Demonstrator:
- The demonstrator will utilize the VIVA eye-tracking system, which operates without traditional cameras, ensuring privacy by not capturing real images of the eyes.
- Multiple identification tasks will be conducted at different time intervals to evaluate the consistency and reliability of person identification based on eye movement data.
Aimed Results:
- Privacy-compliant eye-tracking demonstrator capable of performing person identification without storing real images of the eyes.
- Development of privacy-preserving measures in case identification is achievable under experimental conditions.
- Validation of the VIVA eye-tracking system’s ability to operate effectively while ensuring user privacy.