Light-weight Head Pose Invariant Gaze Tracking

TL;DR

This paper introduces a lightweight, branched CNN architecture for robust, head pose invariant gaze tracking using off-the-shelf cameras, achieving competitive accuracy efficiently.

Contribution

The work proposes a novel branched CNN design and training procedures, including transfer learning and synthetic data use, to enhance gaze estimation robustness without added computational cost.

Findings

Improved robustness to head pose variations.

Achieved competitive accuracy with faster inference.

Effective training strategies using transfer learning and synthetic data.

Abstract

Unconstrained remote gaze tracking using off-the-shelf cameras is a challenging problem. Recently, promising algorithms for appearance-based gaze estimation using convolutional neural networks (CNN) have been proposed. Improving their robustness to various confounding factors including variable head pose, subject identity, illumination and image quality remain open problems. In this work, we study the effect of variable head pose on machine learning regressors trained to estimate gaze direction. We propose a novel branched CNN architecture that improves the robustness of gaze classifiers to variable head pose, without increasing computational cost. We also present various procedures to effectively train our gaze network including transfer learning from the more closely related task of object viewpoint estimation and from a large high-fidelity synthetic gaze dataset, which enable our ten times faster gaze network to achieve competitive accuracy to its current state-of-the-art direct competitor.