Saved in:
Bibliographic Details
Main Authors: Zheng, Henry, Fang, Chenyue, Huang, Rui, Wei, Siyuan, Liu, Xiao, Huang, Gao
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2604.09167
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913021214851072
author Zheng, Henry
Fang, Chenyue
Huang, Rui
Wei, Siyuan
Liu, Xiao
Huang, Gao
author_facet Zheng, Henry
Fang, Chenyue
Huang, Rui
Wei, Siyuan
Liu, Xiao
Huang, Gao
contents Vision-language models (VLMs) have achieved strong performance in multimodal understanding and reasoning, yet grounded reasoning in 3D scenes remains underexplored. Effective 3D reasoning hinges on accurate grounding: to answer open-ended queries, a model must first identify query-relevant objects and regions in a complex scene, and then reason about their spatial and geometric relationships. Recent approaches have demonstrated strong potential for grounded 3D reasoning. However, they often rely on in-domain tuning or hand-crafted reasoning pipelines, which limit their flexibility and zero-shot generalization to novel environments. In this work, we present MAG-3D, a training-free multi-agent framework for grounded 3D reasoning with off-the-shelf VLMs. Instead of relying on task-specific training or fixed reasoning procedures, MAG-3D dynamically coordinates expert agents to address the key challenges of 3D reasoning. Specifically, we propose a planning agent that decomposes the task and orchestrates the overall reasoning process, a grounding agent that performs free-form 3D grounding and relevant frame retrieval from extensive 3D scene observations, and a coding agent that conducts flexible geometric reasoning and explicit verification through executable programs. This multi-agent collaborative design enables flexible training-free 3D grounded reasoning across diverse scenes and achieves state-of-the-art performance on challenging benchmarks.
format Preprint
id arxiv_https___arxiv_org_abs_2604_09167
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle MAG-3D: Multi-Agent Grounded Reasoning for 3D Understanding
Zheng, Henry
Fang, Chenyue
Huang, Rui
Wei, Siyuan
Liu, Xiao
Huang, Gao
Computer Vision and Pattern Recognition
Multiagent Systems
Vision-language models (VLMs) have achieved strong performance in multimodal understanding and reasoning, yet grounded reasoning in 3D scenes remains underexplored. Effective 3D reasoning hinges on accurate grounding: to answer open-ended queries, a model must first identify query-relevant objects and regions in a complex scene, and then reason about their spatial and geometric relationships. Recent approaches have demonstrated strong potential for grounded 3D reasoning. However, they often rely on in-domain tuning or hand-crafted reasoning pipelines, which limit their flexibility and zero-shot generalization to novel environments. In this work, we present MAG-3D, a training-free multi-agent framework for grounded 3D reasoning with off-the-shelf VLMs. Instead of relying on task-specific training or fixed reasoning procedures, MAG-3D dynamically coordinates expert agents to address the key challenges of 3D reasoning. Specifically, we propose a planning agent that decomposes the task and orchestrates the overall reasoning process, a grounding agent that performs free-form 3D grounding and relevant frame retrieval from extensive 3D scene observations, and a coding agent that conducts flexible geometric reasoning and explicit verification through executable programs. This multi-agent collaborative design enables flexible training-free 3D grounded reasoning across diverse scenes and achieves state-of-the-art performance on challenging benchmarks.
title MAG-3D: Multi-Agent Grounded Reasoning for 3D Understanding
topic Computer Vision and Pattern Recognition
Multiagent Systems
url https://arxiv.org/abs/2604.09167