Huiyun Wang, Youjiang Xu, Yahong Han. "Spotting and Aggregating Salient Regions for Video Captioning." Full Paper in ACM MM,2018.

If you find this useful in your work, please consider citing the following reference:

@inproceedings{samwangMM18,
    title = {Spotting and Aggregating Salient Regions for Video Captioning},
    author = {Wang, Huiyun and Xu, Youjiang and Han, Yahong},
    booktitle = {Proceedings of the ACM International Conference on Multimedia (ACM MM)},
    year = {2018}
}

The framework of our method.

The datasets used in the paper are available at the following links:

• Microsoft Video Description Corpus (MSVD): Raw Data Page, Raw Corpus, Processed Data.

• Montreal Video Annotation Description (M-VAD) Dataset: Data Page.

The code is implemented on tensorflow 1.0.1.

The scrips for training and testing of our method can be seen run.sh.

Some examples of our method.