Video-R4: Reinforcing Text-Rich Video Reasoning with Visual Rumination

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Tang, Yolo Y., Shimada, Daiki, Hua, Hang, Huang, Chao, Bi, Jing, Feris, Rogerio, Xu, Chenliang
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918218164076544
author Tang, Yolo Y.
Shimada, Daiki
Hua, Hang
Huang, Chao
Bi, Jing
Feris, Rogerio
Xu, Chenliang
author_facet Tang, Yolo Y.
Shimada, Daiki
Hua, Hang
Huang, Chao
Bi, Jing
Feris, Rogerio
Xu, Chenliang
contents Understanding text-rich videos requires reading small, transient textual cues that often demand repeated inspection. Yet most video QA models rely on single-pass perception over fixed frames, leading to hallucinations and failures on fine-grained evidence. Inspired by how humans pause, zoom, and re-read critical regions, we introduce Video-R4 (Reinforcing Text-Rich Video Reasoning with Visual Rumination), a video reasoning LMM that performs visual rumination: iteratively selecting frames, zooming into informative regions, re-encoding retrieved pixels, and updating its reasoning state. We construct two datasets with executable rumination trajectories: Video-R4-CoT-17k for supervised practice and Video-R4-RL-30k for reinforcement learning. We propose a multi-stage rumination learning framework that progressively finetunes a 7B LMM to learn atomic and mixing visual operations via SFT and GRPO-based RL. Video-R4-7B achieves state-of-the-art results on M4-ViteVQA and further generalizes to multi-page document QA, slides QA, and generic video QA, demonstrating that iterative rumination is an effective paradigm for pixel-grounded multimodal reasoning. Project Page: https://yunlong10.github.io/Video-R4/
format Preprint
id arxiv_https___arxiv_org_abs_2511_17490
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Video-R4: Reinforcing Text-Rich Video Reasoning with Visual Rumination
Tang, Yolo Y.
Shimada, Daiki
Hua, Hang
Huang, Chao
Bi, Jing
Feris, Rogerio
Xu, Chenliang
Computer Vision and Pattern Recognition
Understanding text-rich videos requires reading small, transient textual cues that often demand repeated inspection. Yet most video QA models rely on single-pass perception over fixed frames, leading to hallucinations and failures on fine-grained evidence. Inspired by how humans pause, zoom, and re-read critical regions, we introduce Video-R4 (Reinforcing Text-Rich Video Reasoning with Visual Rumination), a video reasoning LMM that performs visual rumination: iteratively selecting frames, zooming into informative regions, re-encoding retrieved pixels, and updating its reasoning state. We construct two datasets with executable rumination trajectories: Video-R4-CoT-17k for supervised practice and Video-R4-RL-30k for reinforcement learning. We propose a multi-stage rumination learning framework that progressively finetunes a 7B LMM to learn atomic and mixing visual operations via SFT and GRPO-based RL. Video-R4-7B achieves state-of-the-art results on M4-ViteVQA and further generalizes to multi-page document QA, slides QA, and generic video QA, demonstrating that iterative rumination is an effective paradigm for pixel-grounded multimodal reasoning. Project Page: https://yunlong10.github.io/Video-R4/
title Video-R4: Reinforcing Text-Rich Video Reasoning with Visual Rumination
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.17490