# Video-MTR: Reinforced Multi-Turn Reasoning for Long Video Understanding

## TL;DR

Video-MTR introduces an end-to-end reinforced multi-turn reasoning framework that iteratively selects video segments and comprehends questions, significantly improving long video understanding accuracy and efficiency.

## Contribution

It proposes a novel multi-turn reasoning approach with a gated bi-level reward system, enabling end-to-end training without external visual-language models.

## Key findings

- Outperforms existing methods on VideoMME, MLVU, and EgoSchema benchmarks.
- Achieves higher accuracy and efficiency in long video understanding tasks.
- Enables iterative, context-aware video segment selection and question comprehension.

## Abstract

Long-form video understanding, characterized by long-range temporal dependencies and multiple events, remains a challenge. Existing methods often rely on static reasoning or external visual-language models (VLMs), which face issues like complexity and sub-optimal performance due to the lack of end-to-end training. In this paper, we propose Video-MTR, a reinforced multi-turn reasoning framework designed to enable iterative key video segment selection and question comprehension. Unlike traditional video reasoning pipeline, which generate predictions in a single turn, Video-MTR performs reasoning in multiple turns, selecting video segments progressively based on the evolving understanding of previously processed segments and the current question. This iterative process allows for a more refined and contextually aware analysis of the video. To ensure intermediate reasoning process, we introduce a novel gated bi-level reward system, combining trajectory-level rewards based on answer correctness and turn-level rewards emphasizing frame-query relevance. This system optimizes both video segment selection and question comprehension, eliminating the need for external VLMs and allowing end-to-end training. Extensive experiments on benchmarks like VideoMME, MLVU, and EgoSchema demonstrate that Video-MTR outperforms existing methods in both accuracy and efficiency, advancing the state-of-the-art in long video understanding.