Skip to content

ICTMCG/FakeSV

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

20 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

FakeSV

Official repository for "FakeSV: A Multimodal Benchmark with Rich Social Context for Fake News Detection on Short Video Platforms", AAAI 2023. (Please note that the arxiv version is more complete.)

  • Dataset: The video ID (which can be used to infer the video URL) and corresponding annotations have been released. Also, we provide two data split used in the paper, i.e. event-based and temporal.
  • Models: We reproduce some SOTA methods on fake news video detection to provide benchmark results for FakeSV. Codes for our proposed model SV-FEND and other methods are provided.

Environment

Anaconda 4.13.0, python 3.8.5, pytorch 1.10.1 and cuda 11.7. For other libs, please refer to the file requirements.txt.

Application for Data Use

Please sign this agreement and send the signed copy through your institutional email to pengqi.qp@gmail.com.

Data Processing

video-subtitle-extractor

bert-base-chinese

VGG19

C3D

VGGish

MoviepPy

You could use the above repositories to extract features by yourself, or use our pre-extracted features (VGG19/C3D/VGGish). Besides, we also provide five-fold checkpoints for comparison.

Citation

@inproceedings{fakesv, 
title={FakeSV: A Multimodal Benchmark with Rich Social Context for Fake News Detection  on Short Video Platforms}, 
author={Qi, Peng and Bu, Yuyan and Cao, Juan and Ji, Wei and Shui, Ruihao and Xiao,  Junbin and Wang, Danding and Chua, Tat-Seng}, 
booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
year={2023}, 
organization={AAAI} 
} 

Related Survey (ACM Multimedia 2023) and Companion GitHub Repository :

@inproceedings{mvdsurvey, 
title={Combating Online Misinformation Videos: Characterization, Detection, and Future Directions}, 
author={Bu, Yuyan and Sheng, Qiang and Cao, Juan and Qi, Peng and Wang, Danding and Li, Jintao}, 
booktitle={Proceedings of the 31st ACM International Conference on Multimedia}, 
year={2023},
doi={10.1145/3581783.3612426},
publisher = {Association for Computing Machinery},
} 

Related Method Paper (ACL 2023 Findings)

@inproceedings{mvdsurvey, 
title={Two Heads Are Better Than One: Improving Fake News Video Detection by Correlating with Neighbors}, 
author={Qi, Peng  and Zhao, Yuyang  and Shen, Yufeng and Ji, Wei  and Cao, Juan  and Chua, Tat-Seng}, 
booktitle={Findings of the Association for Computational Linguistics: ACL 2023}, 
year={2023},
doi={10.18653/v1/2023.findings-acl.756},
pages = "11947--11959",
publisher = {Association for Computational Linguistics},
} 

About

Official repository for "FakeSV: A Multimodal Benchmark with Rich Social Context for Fake News Detection on Short Video Platforms", AAAI 2023.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages