VideoSSR：影片自監督強化學習

摘要

具可驗證獎勵的強化學習（RLVR）已顯著提升了多模態大型語言模型（MLLMs）的影片理解能力。然而，MLLMs的快速發展正超越現有影片資料集的複雜度，而人工標註高品質新資料的成本依然高昂。本研究探討一個關鍵問題：能否利用影片內在的豐富資訊，自我生成高品質且可驗證的訓練資料？為此，我們引入三項自監督預訓練任務：異常定位、物件計數與時序拼圖。我們建構了影片內在理解基準（VIUBench）以驗證這些任務的難度，結果顯示當前最先進的MLLMs在此類任務上表現明顯不足。基於這些預訓練任務，我們開發了VideoSSR-30K資料集，並提出VideoSSR——一種用於RLVR的新型影片自監督強化學習框架。在涵蓋四大影片領域（通用影片問答、長影片問答、時間定位與複雜推理）的17個基準測試中，廣泛實驗表明VideoSSR能持續提升模型性能，平均改進幅度超過5%。這些成果確立了VideoSSR作為開發更先進MLLMs影片理解能力的強效基礎框架。程式碼已公開於：https://github.com/lcqysl/VideoSSR。

English

Reinforcement Learning with Verifiable Rewards (RLVR) has substantially advanced the video understanding capabilities of Multimodal Large Language Models (MLLMs). However, the rapid progress of MLLMs is outpacing the complexity of existing video datasets, while the manual annotation of new, high-quality data remains prohibitively expensive. This work investigates a pivotal question: Can the rich, intrinsic information within videos be harnessed to self-generate high-quality, verifiable training data? To investigate this, we introduce three self-supervised pretext tasks: Anomaly Grounding, Object Counting, and Temporal Jigsaw. We construct the Video Intrinsic Understanding Benchmark (VIUBench) to validate their difficulty, revealing that current state-of-the-art MLLMs struggle significantly on these tasks. Building upon these pretext tasks, we develop the VideoSSR-30K dataset and propose VideoSSR, a novel video self-supervised reinforcement learning framework for RLVR. Extensive experiments across 17 benchmarks, spanning four major video domains (General Video QA, Long Video QA, Temporal Grounding, and Complex Reasoning), demonstrate that VideoSSR consistently enhances model performance, yielding an average improvement of over 5\%. These results establish VideoSSR as a potent foundational framework for developing more advanced video understanding in MLLMs. The code is available at https://github.com/lcqysl/VideoSSR.

VideoSSR：影片自監督強化學習

VideoSSR: Video Self-Supervised Reinforcement Learning

摘要

Support