VideoRouter: Query-Adaptive Dual Routing for Efficient Long-Video Understanding

2026-05-07 · Kuanwei Lin, Wenhao Zhang, Ge Li arxiv

Video large multimodal models increasingly face a scalability bottleneck: long videos produce excessively long visual-token sequences, which sharply increase memory and latency during inference. While existing compression methods are effective in specific settings, most are either weakly query-aware or apply a fixed compression policy across frames, proving suboptimal when visual evidence is unevenly distributed over time. To address this, we present VideoRouter, a query-adaptive dual-router framework built on InternVL for budgeted evidence allocation. The Semantic Router predicts the dominant allocation policy, choosing between broad temporal coverage and adaptive high-resolution preservation, while the Image Router uses early LLM layers to score frame relevance. This enables aggressive compression on less relevant frames while preserving detail on critical evidence frames. To train both routers, we build Video-QTR-10K for allocation-policy supervision and Video-FLR-200K for frame-relevance supervision. Experiments on VideoMME, MLVU, and LongVideoBench show that VideoRouter consistently improves over the InternVL baseline under comparable or lower budgets, achieving up to a 67.9% token reduction.

📄 PDF Abstract BibTeX arXiv:2605.05848

Code (0)

등록된 구현이 없습니다.

Beyond Frame Selection: Rethinking Long-Video Understanding with MLLMs

2026-08-06 · Ziling Huang, Shin'ichi Satoh 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. …

Routing Before Looking: Query-Adaptive Evidence Acquisition for Long-form Video Understanding

2026-08-21 · Tianyue Wang, Xuying Wu, Yuxiang Ma, Ruiming Liang 외 arxiv

Long-form video understanding remains challenging for video agents due to the mismatch between query demands and evidence acquisition strategies. Although recent planning-before-perception methods outperform query-agnost…

Dynamic Model Routing and Cascading for Efficient LLM Inference: A Survey

2026-02-23 · Yasmin Moslem, John D. Kelleher arxiv

The rapid growth of large language models (LLMs) with diverse capabilities, costs, and domains has created a critical need for intelligent model selection at inference time. While smaller models suffice for routine queri…

Reinforcement Learning

Adaptive LLM Routing under Budget Constraints

2025-08-28 · Pranoy Panda, Raghav Magazine, Chaitanya Devaguptapu, Sho Takemori 외 arxiv

Large Language Models (LLMs) have revolutionized natural language processing, but their varying capabilities and costs pose challenges in practical applications. LLM routing addresses this by dynamically selecting the mo…

Dual-R-DETR: Resolving Query Competition with Pairwise Routing in Transformer Decoders

2025-12-15 · Ye Zhang, Qi Chen, Wenyou Huang, Rui Liu 외 arxiv

Detection Transformers (DETR) formulate object detection as a set prediction problem and enable end-to-end training without post-processing. However, object queries in DETR interact through symmetric self-attention, whic…

Object Detection