Saved in:
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2512.04686 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917130771890176 |
|---|---|
| author | Wang, Yipu Ji, Yuheng Liu, Yuyang Zhou, Enshen Yang, Ziqiang Tian, Yuxuan Qin, Ziheng Liu, Yue Tan, Huajie Chi, Cheng Ma, Zhiyuan Zeng, Daniel Dajun Zheng, Xiaolong |
| author_facet | Wang, Yipu Ji, Yuheng Liu, Yuyang Zhou, Enshen Yang, Ziqiang Tian, Yuxuan Qin, Ziheng Liu, Yue Tan, Huajie Chi, Cheng Ma, Zhiyuan Zeng, Daniel Dajun Zheng, Xiaolong |
| contents | Cross-view correspondence is a fundamental capability for spatial understanding and embodied AI. However, it is still far from being realized in Vision-Language Models (VLMs), especially in achieving precise point-level correspondence, which is crucial for precise affordance interaction. So we propose the Cross-View Point Correspondence (CVPC) task and CrossPoint-Bench, a comprehensive benchmark with hierarchical design, inspired by the human cognitive process of "perceive", "reason", and "correspond". Our evaluation shows the state-of-the-art models (e.g., Gemini-2.5-Pro) still fall far behind humans, with a gap of over 54.65% in overall accuracy, exposing a challenge in transitioning from coarse-grained judgement to fine-grained coordinate prediction. To address this problem, we construct CrossPoint-378K, a dataset with 378K question-answering pairs across 900 scenes, focused on actionable affordance regions that better reflect real-world manipulation and interaction scenarios. Furthermore, we propose CroPond that trained on the CrossPoint-378K dataset. Our CroPond achieves state-of-the-art performance on CrossPoint-Bench, surpassing Gemini-2.5-Pro by 39.7% accuracy, which offers a foundation for advancing future work on cross-view correspondence. The benchmark, dataset, and model are publicly available at https://github.com/WangYipu2002/CrossPoint. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_04686 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Towards Cross-View Point Correspondence in Vision-Language Models Wang, Yipu Ji, Yuheng Liu, Yuyang Zhou, Enshen Yang, Ziqiang Tian, Yuxuan Qin, Ziheng Liu, Yue Tan, Huajie Chi, Cheng Ma, Zhiyuan Zeng, Daniel Dajun Zheng, Xiaolong Computer Vision and Pattern Recognition Cross-view correspondence is a fundamental capability for spatial understanding and embodied AI. However, it is still far from being realized in Vision-Language Models (VLMs), especially in achieving precise point-level correspondence, which is crucial for precise affordance interaction. So we propose the Cross-View Point Correspondence (CVPC) task and CrossPoint-Bench, a comprehensive benchmark with hierarchical design, inspired by the human cognitive process of "perceive", "reason", and "correspond". Our evaluation shows the state-of-the-art models (e.g., Gemini-2.5-Pro) still fall far behind humans, with a gap of over 54.65% in overall accuracy, exposing a challenge in transitioning from coarse-grained judgement to fine-grained coordinate prediction. To address this problem, we construct CrossPoint-378K, a dataset with 378K question-answering pairs across 900 scenes, focused on actionable affordance regions that better reflect real-world manipulation and interaction scenarios. Furthermore, we propose CroPond that trained on the CrossPoint-378K dataset. Our CroPond achieves state-of-the-art performance on CrossPoint-Bench, surpassing Gemini-2.5-Pro by 39.7% accuracy, which offers a foundation for advancing future work on cross-view correspondence. The benchmark, dataset, and model are publicly available at https://github.com/WangYipu2002/CrossPoint. |
| title | Towards Cross-View Point Correspondence in Vision-Language Models |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2512.04686 |