SubsMatch 2.0: Scanpath comparison and classification based on subsequence frequencies.

Area of Science:

• Cognitive Science
• Computer Vision
• Neuroscience

Background:

• Human eye movements balance detailed object focus with environmental awareness.
• Behavioral patterns and action planning are reflected in eye movement sequences.
• Quantifying experimental influences on eye movement data is challenging.

Purpose of the Study:

• To develop an algorithm for extracting sequence-sensitive features from eye movements.
• To classify eye movement sequences based on subsequence frequencies.
• To evaluate the algorithm's performance against state-of-the-art methods.

Main Methods:

• Introduced a novel algorithm for sequence-sensitive feature extraction from eye movements.
• Classified eye movements using the frequencies of small subsequences.
• Evaluated the approach on diverse eye movement datasets from static to dynamic settings.

Main Results:

• The proposed method effectively classifies eye movement sequences across various experimental designs.
• Demonstrated the algorithm's ability to capture sequence-sensitive features.
• Results show superior performance compared to existing methods on rich eye movement data.

Conclusions:

• The developed algorithm provides a robust method for analyzing and classifying eye movement sequences.
• This approach aids in quantifying the influence of experimental factors on behavior.
• The findings advance the understanding of eye movement patterns in different contexts.