Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2605.11723 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911728189571072 |
|---|---|
| author | Wang, Jiyuan Ouyang, Huan Lin, Jiuzhou Lin, Chunyu Fan, Dewen Zhang, Boheng Fan, Haonan Zuo, Fei Sun, Jia Wang, Huaiqing Wang, Honglie Fan, Yiyang Yuan, Zhenlong Li, Zijun Heng, Yongrui Lin, Guosheng Yang, Fan Gao, Tingting |
| author_facet | Wang, Jiyuan Ouyang, Huan Lin, Jiuzhou Lin, Chunyu Fan, Dewen Zhang, Boheng Fan, Haonan Zuo, Fei Sun, Jia Wang, Huaiqing Wang, Honglie Fan, Yiyang Yuan, Zhenlong Li, Zijun Heng, Yongrui Lin, Guosheng Yang, Fan Gao, Tingting |
| contents | In this paper, we propose Concentrate and Concentrate (CaC), a coarse-to-fine anomaly reward model based on Vision-Language Models. During inference, it first conducts a global temporal scan to anchor anomalous time windows, then performs fine-grained spatial grounding within the localized interval, and finally derives robust judgments via structured spatiotemporal Chain-of-Thought reasoning. To equip the model with these capabilities, we construct the first large-scale generated video anomaly dataset with per-frame bounding-box annotations, temporal anomaly windows, and fine-grained attribution labels. Building on this dataset, we design a three-stage progressive training paradigm. The model initially learns spatial and temporal anchoring through single- and multi-frame supervised fine-tuning, and then is optimized by a reinforcement learning strategy based on two-turn Group Relative Policy Optimization (GRPO). Beyond conventional accuracy rewards, we introduce Temporal and Spatial IoU rewards to supervise the intermediate localization process, effectively guiding the model toward more grounded and interpretable spatiotemporal reasoning. Extensive experiments demonstrate that CaC can stably concentrate on subtle anomalies, achieving a 25.7% accuracy improvement on fine-grained anomaly benchmarks and, when used as a reward signal, CaC reduces generated-video anomalies by 11.7% while improving overall video quality. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_11723 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | CaC: Advancing Video Reward Models via Hierarchical Spatiotemporal Concentrating Wang, Jiyuan Ouyang, Huan Lin, Jiuzhou Lin, Chunyu Fan, Dewen Zhang, Boheng Fan, Haonan Zuo, Fei Sun, Jia Wang, Huaiqing Wang, Honglie Fan, Yiyang Yuan, Zhenlong Li, Zijun Heng, Yongrui Lin, Guosheng Yang, Fan Gao, Tingting Computer Vision and Pattern Recognition Artificial Intelligence In this paper, we propose Concentrate and Concentrate (CaC), a coarse-to-fine anomaly reward model based on Vision-Language Models. During inference, it first conducts a global temporal scan to anchor anomalous time windows, then performs fine-grained spatial grounding within the localized interval, and finally derives robust judgments via structured spatiotemporal Chain-of-Thought reasoning. To equip the model with these capabilities, we construct the first large-scale generated video anomaly dataset with per-frame bounding-box annotations, temporal anomaly windows, and fine-grained attribution labels. Building on this dataset, we design a three-stage progressive training paradigm. The model initially learns spatial and temporal anchoring through single- and multi-frame supervised fine-tuning, and then is optimized by a reinforcement learning strategy based on two-turn Group Relative Policy Optimization (GRPO). Beyond conventional accuracy rewards, we introduce Temporal and Spatial IoU rewards to supervise the intermediate localization process, effectively guiding the model toward more grounded and interpretable spatiotemporal reasoning. Extensive experiments demonstrate that CaC can stably concentrate on subtle anomalies, achieving a 25.7% accuracy improvement on fine-grained anomaly benchmarks and, when used as a reward signal, CaC reduces generated-video anomalies by 11.7% while improving overall video quality. |
| title | CaC: Advancing Video Reward Models via Hierarchical Spatiotemporal Concentrating |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2605.11723 |