Skip to content

A Memory Network Approach for Story-based Temporal Summarization of 360° Videos

License

Notifications You must be signed in to change notification settings

sangho-vision/PFMN

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

12 Commits
 
 
 
 
 
 

Repository files navigation

PFMN

main_figure

This project hosts the PyTorch implementation code for our CVPR 2018 paper.

  • Sangho Lee, Jinyoung Sung, Youngjae Yu and Gunhee Kim. A Memory Network Approach for Story-based Temporal Summarization of 360° Videos. In CVPR, 2018. [arxiv]

We propose a novel memory network model named Past-Future Memory Network (PFMN) for story-based temporal summarization of 360° videos. PFMN first computes the score of 81 normal field of view (NFOV) region proposals cropped from the input 360° video, and then recovers a latent, collective storyline using a memory network that involves two external memories to store the embeddings of previously selected subshots and future candidate subshots. Our major contributions are two-fold. First, our work is the first to address story-based temporal summarization of 360° videos. Second, our model is the first attempt to leverage memory networks for video summarization tasks. We empirically validated that the proposed memory network approach outperformed other state-of-the-art methods, not only for view selection but also for story-based temporal summarization in both 360° videos and photostreams.

Acknowledgement

We appreciate Joonil Na, Jaemin Cho and Juyong Kim for helpful comments and discussions.

Authors

Sangho Lee, Jinyoung Sung, Youngjae Yu and Gunhee Kim

Vision and Learning Lab @ Computer Science and Engineering, Seoul National University, Seoul, the Republic of Korea

Reference

If you use this code or dataset as part of any published research, please refer to the following paper.

@inproceedings{lee2018pfmn,
    author    = {Sangho Lee and Jinyoung Sung and Youngjae Yu and Gunhee Kim},
    title     = {{A Memory Network Approach for Story-based Temporal Summarization of 360\deg~Videos}},
    booktitle = {CVPR},
    year      = {2018}
}

License

MIT license

About

A Memory Network Approach for Story-based Temporal Summarization of 360° Videos

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published