Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2504.09130 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908316562620416 |
|---|---|
| author | Wang, Yikun Wang, Siyin Cheng, Qinyuan Fei, Zhaoye Ding, Liang Guo, Qipeng Tao, Dacheng Qiu, Xipeng |
| author_facet | Wang, Yikun Wang, Siyin Cheng, Qinyuan Fei, Zhaoye Ding, Liang Guo, Qipeng Tao, Dacheng Qiu, Xipeng |
| contents | Recent advancements in Large Vision-Language Models have showcased remarkable capabilities. However, they often falter when confronted with complex reasoning tasks that humans typically address through visual aids and deliberate, step-by-step thinking. While existing methods have explored text-based slow thinking or rudimentary visual assistance, they fall short of capturing the intricate, interleaved nature of human visual-verbal reasoning processes. To overcome these limitations and inspired by the mechanisms of slow thinking in human cognition, we introduce VisuoThink, a novel framework that seamlessly integrates visuospatial and linguistic domains. VisuoThink facilitates multimodal slow thinking by enabling progressive visual-textual reasoning and incorporates test-time scaling through look-ahead tree search. Extensive experiments demonstrate that VisuoThink significantly enhances reasoning capabilities via inference-time scaling, even without fine-tuning, achieving state-of-the-art performance in tasks involving geometry and spatial reasoning. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_09130 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | VisuoThink: Empowering LVLM Reasoning with Multimodal Tree Search Wang, Yikun Wang, Siyin Cheng, Qinyuan Fei, Zhaoye Ding, Liang Guo, Qipeng Tao, Dacheng Qiu, Xipeng Computation and Language Recent advancements in Large Vision-Language Models have showcased remarkable capabilities. However, they often falter when confronted with complex reasoning tasks that humans typically address through visual aids and deliberate, step-by-step thinking. While existing methods have explored text-based slow thinking or rudimentary visual assistance, they fall short of capturing the intricate, interleaved nature of human visual-verbal reasoning processes. To overcome these limitations and inspired by the mechanisms of slow thinking in human cognition, we introduce VisuoThink, a novel framework that seamlessly integrates visuospatial and linguistic domains. VisuoThink facilitates multimodal slow thinking by enabling progressive visual-textual reasoning and incorporates test-time scaling through look-ahead tree search. Extensive experiments demonstrate that VisuoThink significantly enhances reasoning capabilities via inference-time scaling, even without fine-tuning, achieving state-of-the-art performance in tasks involving geometry and spatial reasoning. |
| title | VisuoThink: Empowering LVLM Reasoning with Multimodal Tree Search |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2504.09130 |