# Beyond Frame Selection: Rethinking Long-Video Understanding with MLLMs

arXiv: https://arxiv.org/abs/2608.05592 · HTML: https://mlsanity.com/paper/2608.05592/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Multimodal reasoning, one of 17 directions in LLM reasoning.
- Builds on 20 of its 45 references that are in the map.
- First posted on arXiv 6 Aug 2026.

## Placed in

- LLM reasoning / [Multimodal reasoning](https://mlsanity.com/lr/multimodal-reasoning/)
- Multimodal LLMs / vision-language models / [Video understanding with MLLMs](https://mlsanity.com/mllm/video-understanding-with-mllms/)

## Builds on (most cited first, 20 of 20)

- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [GPT-4o System Card](https://mlsanity.com/paper/2410.21276/) (2024)
- [Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context](https://mlsanity.com/paper/2403.05530/) (2024)
- [Qwen3-VL Technical Report](https://mlsanity.com/paper/2511.21631/) (2025)
- [LLaVA-OneVision: Easy Visual Task Transfer](https://mlsanity.com/paper/2408.03326/) (2024)
- [Qwen2 Technical Report](https://mlsanity.com/paper/2407.10671/) (2024)
- [PaLM 2 Technical Report](https://mlsanity.com/paper/2305.10403/) (2023)
- [mPLUG-Owl: Modularization Empowers Large Language Models with Multimodality](https://mlsanity.com/paper/2304.14178/) (2023)
- [VideoLLaMA 2: Advancing Spatial-Temporal Modeling and Audio Understanding in Video-LLMs](https://mlsanity.com/paper/2406.07476/) (2024)
- [LLaVA-Video: Video Instruction Tuning With Synthetic Data](https://mlsanity.com/paper/2410.02713/) (2024)
- [Long Context Transfer from Language to Vision](https://mlsanity.com/paper/2406.16852/) (2024)
- [LongVILA: Scaling Long-Context Visual Language Models for Long Videos](https://mlsanity.com/paper/2408.10188/) (2024)
- [Video-RAG: Visually-aligned Retrieval-Augmented Long Video Comprehension](https://mlsanity.com/paper/2411.13093/) (2024)
- [MiniGPT4-Video: Advancing Multimodal LLMs for Video Understanding with Interleaved Visual-Textual Tokens](https://mlsanity.com/paper/2404.03413/) (2024)
- [Thinking With Videos: Multimodal Tool-Augmented Reinforcement Learning for Long Video Reasoning](https://mlsanity.com/paper/2508.04416/) (2025)
- [VideoMind: A Chain-of-LoRA Agent for Temporal-Grounded Video Reasoning](https://mlsanity.com/paper/2503.13444/) (2025)
- [FOCUS: Efficient Keyframe Selection for Long Video Understanding](https://mlsanity.com/paper/2510.27280/) (2025)
- [Video-MTR: Reinforced Multi-Turn Reasoning for Long Video Understanding](https://mlsanity.com/paper/2508.20478/) (2025)
- [Temporal Chain of Thought: Long-Video Understanding by Thinking in Frames](https://mlsanity.com/paper/2507.02001/) (2025)
- [VideoLucy: Deep Memory Backtracking for Long Video Understanding](https://mlsanity.com/paper/2510.12422/) (2025)

## Abstract (arXiv)

Multimodal Large Language Models (MLLMs) have made strong progress in video understanding, yet long videos remain difficult: the visual token budget grows with video length, so temporally sparse evidence is easily lost. Existing methods, whether frame selection or agent-based exploration, ultimately rely on a single frame set that must trade off broad temporal coverage against fine-grained, question-relevant evidence: mixing both dilutes decisive evidence, while focusing on either sacrifices the other. We propose VideoRouter (VR), which rethinks long-video understanding as coordinating complementary evidence views rather than selecting a single subset of frames. VideoRouter first organizes each video into a question-agnostic temporal hierarchy that partitions it into coarse-to-fine temporally coherent segments. Upper-level nodes capture broad storyline context and event progression, while lower-level nodes preserve fine-grained local details and evidence-bearing moments. This gives rise to two complementary views: a global view for coverage-oriented reasoning and a local view for detail-oriented evidence recovery. We further introduce a verification-guided router that judges which view is better supported by its own selected evidence and decides the final answer. Across six backbones, routing improves over both views in all settings, and the choice of view is shown to be dataset-dependent, confirming that no single evidence granularity is universally preferable. On VideoMME and LongVideoBench, our method outperforms state-of-the-art frame-selection methods by 2.5 and 1.3 points, respectively, under the LLaVA-Video-7B backbone. We will release the code.