VideoAnchor: Reinforcing Subspace-Structured Visual Cues for Coherent Visual-Spatial Reasoning

2025-09-29 · Zhaozhi Wang, Tong Zhang, Mingyue Guo, Yaowei Wang, Qixiang Ye arxiv

Multimodal Large Language Models (MLLMs) have achieved impressive progress in vision-language alignment, yet they remain limited in visual-spatial reasoning. We first identify that this limitation arises from the attention mechanism: visual tokens are overshadowed by language tokens, preventing the model from consistently recognizing the same visual cues across frames. To address this challenge, we draw a novel connection between the self-expressiveness property in sparse subspace clustering and the attention mechanism in Transformers. Building on this insight, we propose VideoAnchor, a plug-and-play module that leverages subspace affinities to reinforce visual cues across frames without retraining, effectively anchoring attention to shared visual structures. Extensive experiments across benchmarks and backbone models show consistent performance gains -- $e.g.$, 3.2% and 4.6% improvements on VSI-Bench and Video-MME (spatial-related tasks) with InternVL2-8B and Qwen2.5VL-72B -- while qualitative analyses demonstrate more coherent subspace partitions and stronger visual grounding. Our codes will be made public available at https://github.com/feufhd/VideoAnchor.

📄 PDF Abstract BibTeX arXiv:2509.25151

Code (0)

등록된 구현이 없습니다.

Tasks

Spatial Reasoning Visual Grounding

Tri-Subspaces Disentanglement for Multimodal Sentiment Analysis

2026-02-23 · Chunlei Meng, Jiabin Luo, Zhenglin Yan, Zhenyu Yu 외 arxiv

Multimodal Sentiment Analysis (MSA) integrates language, visual, and acoustic modalities to infer human sentiment. Most existing methods either focus on globally shared representations or modality-specific features, whil…

Multimodal Intent Recognition Multimodal Sentiment Analysis

CamReasoner: Reinforcing Camera Movement Understanding via Structured Spatial Reasoning

2026-01-30 · Hang Wu, Yujun Cai, Zehao Li, Haonan Ge 외 arxiv

Understanding camera dynamics is a fundamental pillar of video spatial intelligence. However, existing multimodal models predominantly treat this task as a black-box classification, often confusing physically distinct mo…

Binary Classification Spatial Reasoning Visual Reasoning

GraphThinker: Reinforcing Temporally Grounded Video Reasoning with Event Graph Thinking

2026-02-19 · Zixu Cheng, Da Li, Jian Hu, Yuhang Zang 외 arxiv

Video reasoning requires a fine-grained understanding of the temporal dependencies and event-level relations between objects and events in videos. Current Multimodal Large Language Models (MLLMs) are prone to severe temp…

Visual Grounding

Contextual Subspace Manifold Projection for Structural Refinement of Large Language Model Representations

2025-02-12 · Alistair Wren, Beatrice Loxley, Hamish Cadwallader, Simon Beckwith 외

Internal representations within deep neural architectures encode high-dimensional abstractions of linguistic structures, yet they often exhibit inefficiencies in feature distribution, limiting expressiveness and adaptabi…

Language Modeling Language Modelling Large Language Model

MAPS: Multi-Anchor Projection Similarity for Joint Vision-Language Geo-Localization

2026-06-21 · Yutong Hu, Siyuan Tan, Shaocheng Yan, Pengcheng Shi 외 arxiv

Humans localize places by integrating perceptual cues from vision with semantic reasoning from language, forming a scene understanding that is both intuitive and structured. Although existing geo-localization models have…

Scene Understanding