PathGAN: Visual Scanpath Prediction with Generative Adversarial Networks
Abstract
We introduce PathGAN, a deep neural network for visual scanpath prediction trained on adversarial examples. A visual scanpath is defined as the sequence of fixation points over an image defined by a human observer with its gaze. PathGAN is composed of two parts, the generator and the discriminator. Both parts extract features from images using off-the-shelf networks, and train recurrent layers to generate or discriminate scanpaths accordingly. In scanpath prediction, the stochastic nature of the data makes it very difficult to generate realistic predictions using supervised learning strategies, but we adopt adversarial training as a suitable alternative. Our experiments prove how PathGAN improves the state of the art of visual scanpath prediction on the iSUN and Salient360! datasets.
Cite
Text
Assens et al. "PathGAN: Visual Scanpath Prediction with Generative Adversarial Networks." European Conference on Computer Vision Workshops, 2018. doi:10.1007/978-3-030-11021-5_25Markdown
[Assens et al. "PathGAN: Visual Scanpath Prediction with Generative Adversarial Networks." European Conference on Computer Vision Workshops, 2018.](https://mlanthology.org/eccvw/2018/assens2018eccvw-pathgan/) doi:10.1007/978-3-030-11021-5_25BibTeX
@inproceedings{assens2018eccvw-pathgan,
title = {{PathGAN: Visual Scanpath Prediction with Generative Adversarial Networks}},
author = {Assens, Marc and Giró-i-Nieto, Xavier and McGuinness, Kevin and O'Connor, Noel E.},
booktitle = {European Conference on Computer Vision Workshops},
year = {2018},
pages = {406-422},
doi = {10.1007/978-3-030-11021-5_25},
url = {https://mlanthology.org/eccvw/2018/assens2018eccvw-pathgan/}
}