AgentCVR: Active Multi-Agent Cross-Video Reasoning via Script-Simulated Reinforcement Learning
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866913170003591168 |
|---|---|
| author | Qiu, Yilun Wang, Jiahe Yan, Cilin Cai, Jiayin Jiang, Xiaolong Hu, Yao Yuan, Chun |
| author_facet | Qiu, Yilun Wang, Jiahe Yan, Cilin Cai, Jiayin Jiang, Xiaolong Hu, Yao Yuan, Chun |
| contents | Cross-Video Reasoning (CVR) has emerged as a critical frontier in multimodal intelligence, requiring models to retrieve, align, and aggregate evidence distributed across multiple videos. Current Multimodal Large Language Models (MLLMs) often struggle with CVR, as simple single-pass strategies encode multiple videos into a shared compressed context, potentially obscuring rare but critical evidence. In this paper, we propose AgentCVR, a multi-agent framework that treats CVR as an active evidence-acquisition task. AgentCVR employs a Master Agent to iteratively coordinate specialized Visual and Audio Agents for targeted evidence extraction. To ensure efficient training, we introduce Script-Simulated RL, which optimizes the agent's policy with LLM-generated semantic scripts and a lightweight text-based simulator, bypassing costly multimodal inference during online exploration. Experimental results on a comprehensive CVR benchmark show that AgentCVR outperforms single-pass baselines and achieves comparable performance to state-of-the-art closed-source systems, particularly in complex cross-video alignment and localization. To ensure reproducibility, our code is available at https://github.com/wang-jh24/AgentCVR. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_29643 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | AgentCVR: Active Multi-Agent Cross-Video Reasoning via Script-Simulated Reinforcement Learning Qiu, Yilun Wang, Jiahe Yan, Cilin Cai, Jiayin Jiang, Xiaolong Hu, Yao Yuan, Chun Computer Vision and Pattern Recognition Multiagent Systems Cross-Video Reasoning (CVR) has emerged as a critical frontier in multimodal intelligence, requiring models to retrieve, align, and aggregate evidence distributed across multiple videos. Current Multimodal Large Language Models (MLLMs) often struggle with CVR, as simple single-pass strategies encode multiple videos into a shared compressed context, potentially obscuring rare but critical evidence. In this paper, we propose AgentCVR, a multi-agent framework that treats CVR as an active evidence-acquisition task. AgentCVR employs a Master Agent to iteratively coordinate specialized Visual and Audio Agents for targeted evidence extraction. To ensure efficient training, we introduce Script-Simulated RL, which optimizes the agent's policy with LLM-generated semantic scripts and a lightweight text-based simulator, bypassing costly multimodal inference during online exploration. Experimental results on a comprehensive CVR benchmark show that AgentCVR outperforms single-pass baselines and achieves comparable performance to state-of-the-art closed-source systems, particularly in complex cross-video alignment and localization. To ensure reproducibility, our code is available at https://github.com/wang-jh24/AgentCVR. |
| title | AgentCVR: Active Multi-Agent Cross-Video Reasoning via Script-Simulated Reinforcement Learning |
| topic | Computer Vision and Pattern Recognition Multiagent Systems |
| url | https://arxiv.org/abs/2605.29643 |