ADGaze: Anisotropic Gaussian Label Distribution Learning for fine-grained gaze estimation

Open Webpage

Duantengchuan Li, Shutong Wang, Wanli Zhao, Lingyun Kang, Liangshan Dong, Jiazhang Wang, Xiaoguang Wang

Published: 01 Jan 2025, Last Modified: 10 Nov 2025Pattern Recognit. 2025EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•Two essential characteristics revealed: Intra-class similarity and Anisotropy of gaze angles.•We propose multi-loss networks with anisotropic soft label distribution learning to model gaze estimation.•We construct label distributions for different directions using statistical and feature-based similarity approaches.•Experimental results show the effectiveness of the proposed method.