Eye gaze estimation: A survey on deep learning-based approaches

作者：

Highlights：

• Presents an analysis of gaze estimation approaches in unconstrained environments.

• Survey on existing gaze datasets and requirements for a multi-user gaze dataset.

• Analyzes gaze estimation coordinate systems and performance evaluation metrics.

• Analyzes challenges and future directions in gaze estimation.

摘要

•Presents an analysis of gaze estimation approaches in unconstrained environments.•Survey on existing gaze datasets and requirements for a multi-user gaze dataset.•Analyzes gaze estimation coordinate systems and performance evaluation metrics.•Analyzes challenges and future directions in gaze estimation.

论文关键词：Computer vision,Gaze estimation,Deep learning,Eye tracking

论文评审过程：Received 28 August 2021, Revised 15 January 2022, Accepted 11 March 2022, Available online 21 March 2022, Version of Record 29 March 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.116894