Contrastive Gaze

Contrastive gaze research focuses on learning robust representations of eye gaze data, primarily to improve gaze estimation accuracy and efficiency. Current approaches leverage contrastive learning frameworks, often incorporating data augmentation techniques and subject-specific conditioning to enhance model performance, even in unsupervised or low-data settings. This work is significant because it addresses limitations in existing gaze estimation methods, such as the need for large labeled datasets and high computational costs, paving the way for more practical applications in fields like computer-assisted diagnosis and human-computer interaction. Improved gaze estimation has the potential to significantly advance research in cognitive science and enable more intuitive and responsive technologies.