We tackle the problem of predicting saliency maps for videos of dynamic scenes. We note that the accuracy of the maps reconstructed from the gaze data of a fixed number of observers varies with the frame, as it depends on the content of the scene. This issue is particularly pressing when a limited number of observers are available. In such cases, directly minimizing the discrepancy between the predicted and measured saliency maps, as traditional deep-learning methods do, results in overfitting to the noisy data. We propose a noise-aware training (NAT) paradigm that quantifies and accounts for the uncertainty arising from frame-specific gaze data inaccuracy. We show that NAT is especially advantageous when limited training data is available, with experiments across different models, loss functions, and datasets. We also introduce a video game-based saliency dataset, with rich temporal semantics, and multiple gaze attractors per frame. The dataset and source code are available at https://github.com/NVlabs/NAT-saliency.
@article{arxiv.2104.08038,
title = {Noise-Aware Video Saliency Prediction},
author = {Ekta Prashnani and Orazio Gallo and Joohwan Kim and Josef Spjut and Pradeep Sen and Iuri Frosio},
journal= {arXiv preprint arXiv:2104.08038},
year = {2021}
}