Gaze Estimation with an Ensemble of Four Architectures
2021-07-05Code Available1· sign in to hype
Xin Cai, BoYu Chen, Jiabei Zeng, Jiajun Zhang, Yunjia Sun, Xiao Wang, Zhilong Ji, Xiao Liu, Xilin Chen, Shiguang Shan
Code Available — Be the first to reproduce this paper.
ReproduceCode
- github.com/VIPL-TAL-GAZE/GAZE2021pytorch★ 45
Abstract
This paper presents a method for gaze estimation according to face images. We train several gaze estimators adopting four different network architectures, including an architecture designed for gaze estimation (i.e.,iTracker-MHSA) and three originally designed for general computer vision tasks(i.e., BoTNet, HRNet, ResNeSt). Then, we select the best six estimators and ensemble their predictions through a linear combination. The method ranks the first on the leader-board of ETH-XGaze Competition, achieving an average angular error of 3.11^ on the ETH-XGaze test set.