# Appearance-Based Gaze Estimation Using Dilated-Convolutions

## TL;DR

This paper introduces a dilated-convolutional neural network architecture for appearance-based gaze estimation, which preserves spatial resolution to better capture small eye appearance changes and achieves state-of-the-art accuracy.

## Contribution

The novel use of dilated-convolutions in gaze estimation networks enhances feature extraction at high resolution, leading to improved accuracy over existing methods.

## Key findings

- Achieved up to 20.8% accuracy improvement over baseline networks.
- State-of-the-art results on Columbia Gaze and MPIIGaze datasets.
- Effective in cross-subject gaze estimation and eye contact detection.

## Abstract

Appearance-based gaze estimation has attracted more and more attention because of its wide range of applications. The use of deep convolutional neural networks has improved the accuracy significantly. In order to improve the estimation accuracy further, we focus on extracting better features from eye images. Relatively large changes in gaze angles may result in relatively small changes in eye appearance. We argue that current architectures for gaze estimation may not be able to capture such small changes, as they apply multiple pooling layers or other downsampling layers so that the spatial resolution of the high-level layers is reduced significantly. To evaluate whether the use of features extracted at high resolution can benefit gaze estimation, we adopt dilated-convolutions to extract high-level features without reducing spatial resolution. In cross-subject experiments on the Columbia Gaze dataset for eye contact detection and the MPIIGaze dataset for 3D gaze vector regression, the resulting Dilated-Nets achieve significant (up to 20.8%) gains when compared to similar networks without dilated-convolutions. Our proposed Dilated-Net achieves state-of-the-art results on both the Columbia Gaze and the MPIIGaze datasets.