Cross-Encoder for Unsupervised Gaze Representation Learning

Yunjia Sun, Jiabei Zeng, Shiguang Shan, Xilin Chen; Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2021, pp. 3702-3711

In order to train 3D gaze estimators without too many annotations, we propose an unsupervised learning framework, Cross-Encoder, to leverage the unlabeled data to learn suitable representation for gaze estimation. To address the issue that the feature of gaze is always intertwined with the appearance of the eye, Cross-Encoder disentangles the features using a latent-code-swapping mechanism on eye-consistent image pairs and gaze-similar ones. Specifically, each image is encoded as a gaze feature and an eye feature. Cross-Encoder is trained to reconstruct each image in the eye-consistent pair according to its gaze feature and the other's eye feature, but to reconstruct each image in the gaze-similar pair according to its eye feature and the other's gaze feature. Experimental results show the validity of our work. First, using the Cross-Encoder-learned gaze representation, the gaze estimator trained with very few samples outperforms the ones using other unsupervised learning methods, under both within-dataset and cross-dataset protocol. Second, ResNet18 pretrained by Cross-Encoder is competitive with state-of-the-art gaze estimation methods. Third, ablation study shows that Cross-Encoder disentangles the gaze feature and eye feature.