Sparse Evidence Can Suffice: Agentic Evidence Seeking for Multimodal Video Misinformation Detection
2026-07-20 • Computer Vision and Pattern Recognition
Computer Vision and Pattern RecognitionArtificial Intelligence
AI summaryⓘ
The authors address the challenge of finding misinformation in videos, noting that often only a few key clues are important, while most of the video is not helpful. They propose a method called SIEVE, where a smart agent first looks for the important clues and gathers them into a small, useful package. Then, another part checks this smaller set of clues to decide if the video is truthful or not. Their method not only works better than previous ones but also makes it easier to see which clues were used for the decision, making the process clearer and more trustworthy.
multimodal misinformation detectionvideo verificationevidence extractionreinforcement learningmultimodal reasoningevidence-aware traininginformation retrievalagent-based explorationtransparency in AIcompositional evidence
Authors
Haochen Zhao, Yongxiu Xu, Xinkui Lin, Dong Xie, Jiarui Lu, Yuqi Qian, Yubin Wang, Hongbo Xu, Gaopeng Gou
Abstract
Multimodal video misinformation detection is commonly formulated as a holistic video-understanding task, where the entire video and its associated content are processed and judged in a single pass. However, real-world misinformation often exhibits a sparse and compositional evidence structure: a reliable decision may depend on only a few coupled clues, while most video content contributes limited additional information. Exhaustive multimodal reasoning may therefore introduce substantial redundancy and obscure decisive evidence. This motivates decoupling evidence acquisition from verification: first identifying sparse, decision-relevant clues and then judging veracity based on the acquired evidence. Accordingly, we propose SIEVE, a framework for Sparse Interactive Evidence Verification via Extraction in multimodal video misinformation detection. An evidence-seeking agent actively explores the available multimodal evidence and constructs a compact evidence package, which is then used by a verifier to determine veracity. The agent is trained with supervised evidence-seeking trajectories and an evidence-aware reinforcement learning objective that promotes informative evidence acquisition while discouraging unnecessary or invalid interactions. Experiments on multiple video misinformation benchmarks show that SIEVE consistently outperforms the evaluated baselines and supports reliable verification using compact evidence packages. Moreover, the resulting acquisition process provides an explicit and inspectable evidence trail, improving the transparency and groundedness of multimodal misinformation detection.