Guardado en:
Detalles Bibliográficos
Autores principales: Gao, Xiangjun, Zhang, Zhensong, Chen, Dave Zhenyu, Xu, Songcen, Quan, Long, Pérez-Pellitero, Eduardo, Jang, Youngkyoon
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:https://arxiv.org/abs/2601.11442
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866911380427243520
author Gao, Xiangjun
Zhang, Zhensong
Chen, Dave Zhenyu
Xu, Songcen
Quan, Long
Pérez-Pellitero, Eduardo
Jang, Youngkyoon
author_facet Gao, Xiangjun
Zhang, Zhensong
Chen, Dave Zhenyu
Xu, Songcen
Quan, Long
Pérez-Pellitero, Eduardo
Jang, Youngkyoon
contents We propose Map2Thought, a framework that enables explicit and interpretable spatial reasoning for 3D VLMs. The framework is grounded in two key components: Metric Cognitive Map (Metric-CogMap) and Cognitive Chain-of-Thought (Cog-CoT). Metric-CogMap provides a unified spatial representation by integrating a discrete grid for relational reasoning with a continuous, metric-scale representation for precise geometric understanding. Building upon the Metric-CogMap, Cog-CoT performs explicit geometric reasoning through deterministic operations, including vector operations, bounding-box distances, and occlusion-aware appearance order cues, producing interpretable inference traces grounded in 3D structure. Experimental results show that Map2Thought enables explainable 3D understanding, achieving 59.9% accuracy using only half the supervision, closely matching the 60.9% baseline trained with the full dataset. It consistently outperforms state-of-the-art methods by 5.3%, 4.8%, and 4.0% under 10%, 25%, and 50% training subsets, respectively, on the VSI-Bench.
format Preprint
id arxiv_https___arxiv_org_abs_2601_11442
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Map2Thought: Explicit 3D Spatial Reasoning via Metric Cognitive Maps
Gao, Xiangjun
Zhang, Zhensong
Chen, Dave Zhenyu
Xu, Songcen
Quan, Long
Pérez-Pellitero, Eduardo
Jang, Youngkyoon
Computer Vision and Pattern Recognition
Artificial Intelligence
We propose Map2Thought, a framework that enables explicit and interpretable spatial reasoning for 3D VLMs. The framework is grounded in two key components: Metric Cognitive Map (Metric-CogMap) and Cognitive Chain-of-Thought (Cog-CoT). Metric-CogMap provides a unified spatial representation by integrating a discrete grid for relational reasoning with a continuous, metric-scale representation for precise geometric understanding. Building upon the Metric-CogMap, Cog-CoT performs explicit geometric reasoning through deterministic operations, including vector operations, bounding-box distances, and occlusion-aware appearance order cues, producing interpretable inference traces grounded in 3D structure. Experimental results show that Map2Thought enables explainable 3D understanding, achieving 59.9% accuracy using only half the supervision, closely matching the 60.9% baseline trained with the full dataset. It consistently outperforms state-of-the-art methods by 5.3%, 4.8%, and 4.0% under 10%, 25%, and 50% training subsets, respectively, on the VSI-Bench.
title Map2Thought: Explicit 3D Spatial Reasoning via Metric Cognitive Maps
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2601.11442