/AFF-Net

Adaptive Feature Fusion Network for Gaze Tracking in Mobile Tablets.

Primary LanguagePython

ReadMe

Codes for ICPR2020 paper Adaptive Feature Fusion Network for Gaze Tracking in Mobile Tablets.

The project is based on:

numpy==1.20.1
opencv-python==4.2.0.34
torch==1.5.0
torchvision==0.6.0

As we conduct experiments on re-face-detected GazeCapture dataset, the data-loader codes are not universal with original dataset. For details and codes about data processings, SOTA methods and benchmark, please refer to our survey paper "Appearance-based Gaze Estimation With Deep Learning: A Review and Benchmark".