Elamur Research⌕
☰
home / papers / 2026 / Perception, Understanding and Reasoning: A Multimodal Benchmark for Video Fake News Detection

Perception, Understanding and Reasoning: A Multimodal Benchmark for Video Fake News Detection

Paper · 2026 · Cui Yakun, Peng Qi, Fushuo Huo et al.

Summary

Most benchmarks for video fake news detection (VFND) score only the final real/fake label. They cannot show whether a multimodal large language model (MLLM) actually saw the right evidence, understood the story, or reasoned soundly. The authors address this with POVFNDB, a process-oriented benchmark. It is built on 3,624 videos from the FakeSV dataset and adds about 36,240 human-checked question–answer pairs, organized into 10 sub-tasks and 15 metrics. These cover three stages: perception, understanding, and reasoning. Perception tasks also separate creator-added content (overlaid text, edits) from original shooting footage. Tests of open-source and proprietary MLLMs show clear weaknesses. Even the strongest model, Gemini2.5-Flash, performs poorly on most perception and reasoning sub-tasks. The authors also propose POVFND-CoT, a chain-of-thought framework that uses these process signals to produce reasoning data. Fine-tuning Qwen2.5-VL-7B-Instruct on this data gives state-of-the-art detection results. The work suggests that reliable, explainable VFND depends on domain-specific reasoning training, not only on general model ability.

Key points

  • POVFNDB (Process-oriented Video Fake News Detection Benchmark): 36,240 QA items over 3,624 FakeSV videos, with 10 tasks and 15 metrics. Example tasks are key element grounding (IoU), shooting angle counting, creator-added text color/position, human identity recognition, and news type/theme understanding.
  • Reasoning metrics: the final detection task is scored on four measures. These are entity hit rate (ENHR), external-knowledge hit rate (KHR), rationale hit rate (RHR), and accuracy.
  • Annotation pipeline: Gemini2.5-Flash drafted the initial answers and extracted entities, knowledge, and rationales. o1 then checked consistency and plausibility, followed by manual review.
  • CoT variants: the authors tested CAC-, OSF-, and ALL-POVFND-CoT. When used only for prompting, ALL-POVFND-CoT scored lower than OSF-POVFND-CoT, which suggests MLLMs struggle to weigh mixed evidence types.
  • Strong baseline: Qwen2.5-VL-7B-Instruct was fine-tuned on ALL-POVFND-CoT reasoning traces generated by Gemini2.5-Flash. It reached state-of-the-art accuracy on the test split.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓Peking UniversityBeijing · China244 works · 39,359 citations🎓National University of SingaporeSingapore · Singapore180 works · 34,304 citations🎓Hong Kong University of Science and TechnologyHong Kong · China142 works · 25,136 citations🎓Beijing University of Posts and TelecommunicationsBeijing · China57 works · 32,843 citations🎓Southeast UniversityNanjing · China28 works · 354 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksComputer Vision and Pattern Recognition · 2,869 worksSocial Sciences · 1,772 worksSociology and Political Science · 577 works

🧭 Topics 3Research subjects and languages the work is about.Reasoning · 1,426 worksEvaluation & benchmarks · 1,220 worksVision & language · 657 works

Cite

Cite as Cui Yakun, Peng Qi, Fushuo Huo, Hang Du, Weijie Shi, Juntao Dai, Zhenghao Zhu and Sirui Han. 2026. Perception, Understanding and Reasoning: A Multimodal Benchmark for Video Fake News Detection. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.acl-long.2103

@inproceedings{yakun2026perception,
  title = {{Perception, Understanding and Reasoning: A Multimodal Benchmark for Video Fake News Detection}},
  author = {Cui Yakun, Peng Qi, Fushuo Huo, Hang Du, Weijie Shi, Juntao Dai, Zhenghao Zhu and Sirui Han},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.2103},
  url = {https://doi.org/10.18653/v1/2026.acl-long.2103}
}

References

39 references, 6 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Hongbo Bai, Yujin Zhou, Yile Wu, Chi-Min Chan, Pengcheng Wen, Kunhao Pan, Sirui Han, and Yike Guo. 2026. Glance-or-gaze: Incentivizing lmms to adaptively focus search via reinforcement learning. arXiv preprint arXiv:2601.13942. find
  2. Mary Angela Bock, Robert J. Richardson, Christopher T. Assaf, and Dariya Tsyrenzhapova. 2023. Improvisation, economy, and mtv moves: Online news and video production style. Electronic News, 17(3):146– 163. find
  3. Yuyan Bu, Qiang Sheng, Juan Cao, Peng Qi, Danding Wang, and Jintao Li. 2024. Fakingrecipe: Detecting fake news on short video platforms from the perspective of creative process. In Proceedings of the 32nd ACM International Conference on Multimedia, pages 1351–1360. find
  4. Fabian Caba Heilbron, Victor Escorcia, Bernard Ghanem, and Juan Carlos Niebles. 2015. Activitynet: A large-scale video benchmark for human activity understanding. In Proceedings of the ieee conference on computer vision and pattern recognition, pages 961–970. find
  5. Xiaowei Chi, Chun-Kai Fan, Hengyuan Zhang, Xingqun Qi, Rongyu Zhang, Anthony Chen, Chi-min Chan, Wei Xue, Qifeng Liu, Shanghang Zhang, and 1 others. 2024. Eva: An embodied world model for future video anticipation. arXiv preprint arXiv:2410.15461. — [arXiv:2410.15461](https://arxiv.org/abs/2410.15461) find
  6. Arianna D’ulizia, Maria Chiara Caschera, Fernando Ferri, and Patrizia Grifoni. 2021. Fake news detection: a survey of evaluation datasets. PeerJ Computer Science, 7:e518. find
  7. Hao Fei, Shengqiong Wu, Wei Ji, Hanwang Zhang, Meishan Zhang, Mong-Li Lee, and Wynne Hsu. 2024. Video-of-thought: Step-by-step video reasoning from perception to cognition. arXiv preprint arXiv:2501.03230. — [arXiv:2501.03230](https://arxiv.org/abs/2501.03230) find
  8. Kaituo Feng, Kaixiong Gong, Bohao Li, Zonghao Guo, Yibing Wang, Tianshuo Peng, Junfei Wu, Xiaoying Zhang, Benyou Wang, and Xiangyu Yue. 2025. Video-r1: Reinforcing video reasoning in mllms. arXiv preprint arXiv:2503.21776. — [arXiv:2503.21776](https://arxiv.org/abs/2503.21776) find
  9. Hui Han, Siyuan Li, Jiaqi Chen, Yiwen Yuan, Yuling Wu, Yufan Deng, Chak Tou Leong, Hanwen Du, Junchen Fu, Youhua Li, and 1 others. 2025. Video-bench: Human-aligned video generation benchmark. In Proceedings of the Computer Vision and Pattern Recognition Conference, pages 18858–18868. find
  10. Ruiyan Han, Zhen Fang, XinYu Sun, Yuchen Ma, Ziheng Wang, Yu Zeng, Zehui Chen, Lin Chen, Wenxuan Huang, Wei-Jie Xu, and 1 others. 2026. Unicorn: Towards self-improving unified multimodal models through self-generated supervision. arXiv preprint arXiv:2601.03193.
  11. Ye Liu, Zongyang Ma, Zhongang Qi, Yang Wu, Ying Shan, and Chang W Chen. 2024. Et bench: Towards open-ended event-level video-language understanding. Advances in Neural Information Processing Systems, 37:32076–32110. find
  12. Yue Ma, Yulong Liu, Qiyuan Zhu, Xiangpeng Yang, Kunyu Feng, Xinhua Zhang, Zexuan Yan, Zhifeng Li, Sirui Han, Chenyang Qi, and 1 others. Effivmt: Video motion transfer via efficient spatial-temporal decoupled finetuning. In The Fourteenth International Conference on Learning Representations. find
  13. Qiong Nan, Qiang Sheng, Juan Cao, Beizhe Hu, Danding Wang, and Jintao Li. 2024. Let silence speak: Enhancing fake news detection with generated comments from large language models. In Proceedings of the 33rd ACM International Conference on Information and Knowledge Management, pages 1732–1742. find
  14. Munan Ning, Bin Zhu, Yujia Xie, Bin Lin, Jiaxi Cui, Lu Yuan, Dongdong Chen, and Li Yuan. 2023. Video-bench: A comprehensive benchmark and toolkit for evaluating video-based large language models. arXiv preprint arXiv:2311.16103. — [arXiv:2311.16103](https://arxiv.org/abs/2311.16103) find
  15. Andrea Pereira, Elizabeth Harris, and Jay J. Van Bavel. 2023. Identity concerns drive belief: The impact of partisan identity on the belief and dissemination of true and false news. Group Processes & Intergroup Relations, 26(1):24–47. find
  16. Peng Qi, Yuyan Bu, Juan Cao, Wei Ji, Ruihao Shui, Junbin Xiao, Danding Wang, and Tat-Seng Chua. 2023a. Fakesv: A multimodal benchmark with rich social context for fake news detection on short video platforms. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, pages 14444– 1… find
  17. Peng Qi, Zehong Yan, Wynne Hsu, and Mong Li Lee. 2024. Sniffer: Multimodal large language model for explainable out-of-context misinformation detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 13052–13062. find
  18. Peng Qi, Yuyang Zhao, Yufeng Shen, Wei Ji, Juan Cao, and Tat-Seng Chua. 2023b. Two heads are better than one: Improving fake news video detection by correlating with neighbors. arXiv preprint arXiv:2306.05241. find
  19. Xingqun Qi, Yatian Wang, Hengyuan Zhang, Jiahao Pan, Wei Xue, Shanghang Zhang, Wenhan Luo, Qifeng Liu, and Yike Guo. 2025. CoΘ {3} gesture: Towards coherent concurrent co-speech 3d gesture generation with interactive diffusion. arXiv preprint arXiv:2505.01746. — [arXiv:2505.01746](https://arxiv.org/abs/2505.01746) find
  20. Long Qian, Juncheng Li, Yu Wu, Yaobo Ye, Hao Fei, Tat-Seng Chua, Yueting Zhuang, and Siliang Tang. 2024a. Momentor: Advancing video large language model with fine-grained temporal reasoning. arXiv preprint arXiv:2402.11435. — [arXiv:2402.11435](https://arxiv.org/abs/2402.11435) find
  21. Rui Qian, Xiaoyi Dong, Pan Zhang, Yuhang Zang, Shuangrui Ding, Dahua Lin, and Jiaqi Wang. 2024b. Streaming long video understanding with large language models. Advances in Neural Information Processing Systems, 37:119336–119360. find
  22. Jing Shen, Yanjia Wang, Shengze Wang, Yuping Zhang, and Haibo Liu. 2025. Multi-modal similarity guided adaptive fusion network for short video fake news detection. In Proceedings of the 2025 International Conference on Multimedia Retrieval, pages 1145– 1153. find
  23. Shangkun Sun, Xiaoyu Liang, Songlin Fan, Wenxu Gao, and Wei Gao. 2025. Ve-bench: Subjective-aligned benchmark suite for text-driven video editing quality assessment. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 39, pages 7105– 7113. find
  24. Yunlong Tang, Jing Bi, Siting Xu, Luchuan Song, Susan Liang, Teng Wang, Daoan Zhang, Jie An, Jingyang Lin, Rongyi Zhu, and 1 others. 2025a. Video understanding with large language models: A survey. IEEE Transactions on Circuits and Systems for Video Technology.
  25. Yunlong Tang, Gen Zhan, Li Yang, Yiting Liao, and Chenliang Xu. 2025b. Cardiff: Video salient object ranking chain of thought reasoning for saliency prediction with diffusion. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 39, pages 7302–7310. find
  26. Junxi Wang, Yaxiong Wang, Lechao Cheng, and Zhun Zhong. 2025a. Fakesv-vlm: Taming vlm for detecting fake short-video news via progressive mixture-ofexperts adapter. arXiv preprint arXiv:2508.19639. — [arXiv:2508.19639](https://arxiv.org/abs/2508.19639) find
  27. Xiaohan Wang, Yuhui Zhang, Orr Zohar, and Serena Yeung-Levy. 2024. Videoagent: Long-form video understanding with large language model as agent. In European Conference on Computer Vision, pages 58–76. Springer. find
  28. Yihao Wang, Zhong Qian, and Peifeng Li. 2025b. Fmnv: A dataset of media-published news videos for fake news detection. In International Conference on Intelligent Computing, pages 321–332. Springer. find
  29. Cheng Xu and Nan Yan. 2025. Triplefact: Defending data contamination in the evaluation of llm-driven fake news detection. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 8808– 8823.
  30. Zhi Zeng, Jiaying Wu, Minnan Luo, Herun Wan, Xiangzheng Kong, Zihan Ma, Guang Dai, and Qinghua Zheng. 2025. Imol: Incomplete-modality-tolerant learning for multi-domain fake news video detection. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Lo…
  31. Litian Zhang, Xiaoming Zhang, Ziyi Zhou, Xi Zhang, Philip S Yu, and Chaozhuo Li. 2025a. Knowledge-aware multimodal pre-training for fake news detection. Information Fusion, 114:102715. find
  32. Zicheng Zhang, Ziheng Jia, Haoning Wu, Chunyi Li, Zijian Chen, Yingjie Zhou, Wei Sun, Xiaohong Liu, Xiongkuo Min, Weisi Lin, and 1 others. 2025b. Q-bench-video: Benchmark the video quality understanding of lmms. In Proceedings of the Computer Vision and Pattern Recognition Conference, pages 3229–323… find
  33. Jianlong Zhou, Xianglin Miao, Feijuan He, and Yalin Miao. 2022. Effects of font style and font color in news text on user cognitive load in intelligent user interfaces. IEEE Access, 10:10719–10730. find
  34. Xinyi Zhou and Reza Zafarani. 2019. Network-based fake news detection: A pattern-driven approach. ACM SIGKDD explorations newsletter, 21(2):48–60. find
  35. Yujin Zhou, Pengcheng Wen, Jiale Chen, Boqin Yin, Han Zhu, Jiaming Ji, Juntao Dai, Chi-Min Chan, and Sirui Han. 2026. What, whether and how? unveiling process reward models for thinking with images reasoning. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 40, pages 29071– 2… find
  36. Han Zhu, Juntao Dai, Jiaming Ji, Haoran Li, Chengkun Cai, Pengcheng Wen, Chi-Min Chan, Boyuan Chen, Yaodong Yang, Sirui Han, and Yike Guo. 2025. Safemt: Multi-turn safety for multimodal language models. Preprint, arXiv:2510.12133.
  37. Linlin Zong, Wenmin Lin, Jiahui Zhou, Xinyue Liu, Xianchao Zhang, Bo Xu, and Shimin Wu. 2025. Text-guided fine-grained counterfactual inference for short video fake news detection. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 39, pages 1237–1245. find
  38. Linlin Zong, Jiahui Zhou, Wenmin Lin, Xinyue Liu, Xianchao Zhang, and Bo Xu. 2024. Unveiling opinion evolution via prompting and diffusion for short video fake news detection. In Findings of the Association for Computational Linguistics ACL 2024, pages 10817–10826. find
  39. Xiang Yue, Yuansheng Ni, Kai Zhang, Tianyu Zheng, Ruoqi Liu, Ge Zhang, Samuel Stevens, Dongfu Jiang, Weiming Ren, Yuxuan Sun, and 1 others. 2024. Mmmu: A massive multi-discipline multimodal understanding and reasoning benchmark for expert agi. In Proceedings of the IEEE/CVF Conference on Computer Vi…

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/yakun-2026-perception-understanding-reasoning-multimodal-benchmark/