Few-Shot Adaptive Gaze Estimation

Seonwook Park, Shalini De Mello, Pavlo Molchanov, Umar Iqbal, Otmar Hilliges, Jan Kautz; The IEEE International Conference on Computer Vision (ICCV), 2019, pp. 9368-9377

Abstract


Inter-personal anatomical differences limit the accuracy of person-independent gaze estimation networks. Yet there is a need to lower gaze errors further to enable applications requiring higher quality. Further gains can be achieved by personalizing gaze networks, ideally with few calibration samples. However, over-parameterized neural networks are not amenable to learning from few examples as they can quickly over-fit. We embrace these challenges and propose a novel framework for Few-shot Adaptive GaZE Estimation (Faze) for learning person-specific gaze networks with very few (<= 9) calibration samples. Faze learns a rotation-aware latent representation of gaze via a disentangling encoder-decoder architecture along with a highly adaptable gaze estimator trained using meta-learning. It is capable of adapting to any new person to yield significant performance gains with as few as 3 samples, yielding state-of-the-art performance of 3.18-deg on GazeCapture, a 19% improvement over prior art. We open-source our code at https://github.com/NVlabs/few_shot_gaze

Related Material


[pdf] [supp]
[bibtex]
@InProceedings{Park_2019_ICCV,
author = {Park, Seonwook and Mello, Shalini De and Molchanov, Pavlo and Iqbal, Umar and Hilliges, Otmar and Kautz, Jan},
title = {Few-Shot Adaptive Gaze Estimation},
booktitle = {The IEEE International Conference on Computer Vision (ICCV)},
month = {October},
year = {2019}
}