让视频模型分步思考，提升推理能力。
VIDEOP2R: Video Understanding from Perception to Reasoning
- 将感知与推理拆解为独立步骤建模，分阶段优化。
- 在7个基准中6个达到顶尖性能，感知输出已足够支持推理。
- 适合需要精准视频理解与逻辑推理的科研与工业场景。
强化学习微调（RFT）是一种两阶段框架，包含监督微调（SFT）和强化学习（RL），已被证明能有效提升大语言模型（LLM）的推理能力。然而，将其扩展至大视频语言模型（LVLM）仍面临挑战。本文提出 VideoP2R，一种面向过程的视频RFT框架，通过将感知与推理建模为独立过程来增强视频推理能力。在SFT阶段，设计三步流程生成 VideoP2R-CoT-162K——一个高质量、面向过程的思维链（CoT）数据集，用于感知与推理。在RL阶段，引入新型过程感知组相对策略优化（PA-GRPO）算法，分别对感知与推理提供奖励。大量实验表明，VideoP2R在7个视频推理与理解基准中，有6个达到当前最优性能。消融实验证实了过程感知建模与PA-GRPO的有效性，并表明模型的感知输出已具备足够的信息支持下游推理。项目主页见：https://videop2r.github.io/videop2r/。
原文摘要 · Abstract (English)
Reinforcement fine-tuning (RFT), a two-stage framework consisting of supervised fine-tuning (SFT) and reinforcement learning (RL) has shown promising results on improving reasoning ability of large language models (LLMs). Yet extending RFT to large video language models (LVLMs) remains challenging. We propose VideoP2R, a novel process-aware video RFT framework that enhances video reasoning by modeling perception and reasoning as distinct processes. In the SFT stage, we develop a three-step pipeline to generate VideoP2R-CoT-162K, a high-quality, process-aware chain-of-thought (CoT) dataset for perception and reasoning. In the RL stage, we introduce a novel process-aware group relative policy optimization (PA-GRPO) algorithm that supplies separate rewards for perception and reasoning. Extensive experiments show that VideoP2R achieves state-of-the-art (SotA) performance on six out of seven video reasoning and understanding benchmarks. Ablation studies further confirm the effectiveness of our process-aware modeling and PA-GRPO and demonstrate that model's perception output is information-sufficient for downstream reasoning. Our project page is available at https://videop2r.github.io/videop2r/.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。