CasualGaze: Towards Modeling and Recognizing Casual Gaze Behavior for Efficient Gaze-based Object Selection

• DOI: 10.48550/arxiv.2408.12710
• Publication Date: 2024-08-22
• Authors: Yingtian Shi; Yukang Yan; Zisu Li; Liang Chen; Yuntao Wang; Chun Yu; Yuanchun Shi
• Keywords: Casual

We present CasualGaze, a novel eye-gaze-based target selection technique to support natural and casual eye-gaze input. Unlike existing solutions that require users keep the center on actively, CasualGaze allows glance at object complete simply. To understand gaze behavior, we studied spatial distribution of for different layouts user behavior in simulated real-world environment. Results revealed impacts parameters, speed randomness features gaze, special patterns "blurred areas". Based results, devised algorithms, employing bivariate Gaussian model along with temporal compensation voting algorithms robust prediction. Usability evaluation study showed significant improvements recognition compared two baseline techniques. Subjective ratings comments further supported preference regarding efficiency, accuracy, stability.