BROWSE
Detail
Conference Paper
SpatialAgent: Spatial Question Answering with LLM Agent and Perception Models
Cited - time in 
Share 
Share
- Authors
- Hsiang-Wei Huang, Junbin Lu, Pyongkun Kim, Jianxu Shangguan, Jen-Hao Cheng, Kuang-Ming Chen, Cheng-Yen Yang, Bahaa Alattar, Yi-Ru Lin, Sangwon Kim, Kwangju Kim, Chung-I Huang, Jenq-Neng Hwang
- Issue Date
- 2026-08
- Citation
- International Conference on Multimedia Information Processing and Retrieval (MIPR) 2026, pp.1-7
- Publisher
- IEEE
- Language
- English
- Type
- Conference Paper
- Abstract
- Spatial understanding has been a challenging task for existing Multi-modal Large Language Models (MLLMs). Previous methods leverage large-scale MLLM finetuning to enhance MLLM’s spatial understanding ability. In this paper, we present a data-efficient approach. We propose an LLM agent system with strong and advanced spatial reasoning ability, which can be used to solve the challenging spatial question answering task in complex indoor warehouse scenarios. Our system integrates multiple tools that allow the LLM agent to conduct spatial reasoning and API tool interaction to answer the given complicated spatial question. Extensive evaluations demonstrate that our system achieves high accuracy and efficiency in tasks such as object retrieval, counting, and distance estimation. On the 2025 AI City Challenge Physical AI Spatial Intelligence Warehouse dataset, our system achieves 95.86% accuracy, ranking 1st place among all teams. On the SpatialBench benchmark, our system further attains 84.2% accuracy, setting a new state-of-the-art for zero-shot spatial question answering. The code is available at https://github.com/hsiangwei0903/SpatialAgent.
- Keyword
- LLM Agent, Spatial Question Answering
- KSP Keywords
- Accuracy and efficiency, Efficient approach, Language Models, Multi-modal, Object Retrieval, Question Answering, Reasoning ability, Spatial intelligence, Spatial reasoning, Zero-shot, agent system