Gaze Sequence

Gaze sequence research focuses on understanding and modeling the patterns of eye movements, aiming to predict and generate realistic gaze behavior. Current research employs deep learning models, including diffusion models and transformer-based architectures, to analyze and synthesize gaze sequences from various image types, such as 2D and 360° images, and even to predict gaze from other modalities like radiology reports. This work is significant for advancing human-computer interaction, improving AI systems, and providing insights into cognitive processes, particularly through the analysis of gaze events like fixations and saccades and their contribution to model predictions.