Awesome
This repository provides the code in our ECCV paper
" Learning to Predict Salient Faces: A Novel Visual-Audio Saliency Model "
Abstract
Recently, video streams have occupied a large proportion of Internet traffic, most of which contain human faces. Hence, it is necessary to predict saliency on multiple-face videos, which can provide attention cues for many content based applications. However, most of multiple-face saliency prediction works only consider visual information and ignore audio, which is not consistent with the naturalistic scenarios. Several behavioral studies have established that sound influences human attention, especially during the speech turn-taking in multipleface videos. In this paper, we thoroughly investigate such influences by establishing a large-scale eye-tracking database of Multiple-face Video in Visual-Audio condition (MVVA). Inspired by the findings of our investigation, we propose a novel multi-modal video saliency model consisting of three branches: visual, audio and face. The visual branch takes the RGB frames as the input and encodes them into visual feature maps. The audio and face branches encode the audio signal and multiple cropped faces, respectively. A fusion module is introduced to integrate the information from three modalities, and to generate the final saliency map. Experimental results show that the proposed method outperforms 11 state-of-the-art saliency prediction works. It performs closer to human multi-modal attention.
Network
Requirements
- python 3.7
- pytorch 1.1.0
- opencv
- librosa
The dependencies can be installed through requirements.txt
Inference
Download the pretrained model from here, the gmm map generated by face branch and our MVVA database from here, and run the demo inference code
python main.py
Checklist
- Update the Visual branch, Audio branch and Fusion module
- Update the Face branch
Citation
If you find this repository helpful, you may cite:
@article{liu2020visualaudio,
title={Learning to Predict Salient Faces: A Novel Audio-Visual Saliency Model},
author={Yufan Liu; Minglang Qiao; Mai Xu; Bing Li; Weiming Hu; Ali Borji},
booktitle=={Proceedings of the european conference on computer vision (eccv)},
year={2020}
}