SpatialReasoner: Active Perception for Large-Scale 3D Scene Understanding
By: Hongpei Zheng , Shijie Li , Yanran Li and more
Potential Business Impact:
Helps robots understand whole houses, not just rooms.
Spatial reasoning in large-scale 3D environments remains challenging for current vision-language models, which are typically constrained to room-scale scenarios. We introduce H$^2$U3D (Holistic House Understanding in 3D), a 3D visual question answering dataset designed for house-scale scene understanding. H$^2$U3D features multi-floor environments spanning up to three floors and 10-20 rooms, covering more than 300 m$^2$. Through an automated annotation pipeline, it constructs hierarchical coarse-to-fine visual representations and generates diverse question-answer pairs with chain-of-thought annotations. We further propose SpatialReasoner, an active perception framework that autonomously invokes spatial tools to explore 3D scenes based on textual queries. SpatialReasoner is trained through a two-stage strategy: a supervised cold start followed by reinforcement learning with an adaptive exploration reward that promotes efficient exploration while discouraging redundant operations. Extensive experiments demonstrate that SpatialReasoner achieves state-of-the-art performance on H$^2$U3D, outperforming strong baselines including GPT-4o and Gemini-2.5-Pro. Notably, our method attains superior results while using only 3-4 images in total on average, compared to baselines requiring 16+ images, highlighting the effectiveness of our coarse-to-fine active exploration paradigm.
Similar Papers
SpatialReasoner: Towards Explicit and Generalizable 3D Spatial Reasoning
CV and Pattern Recognition
Helps computers understand 3D space like humans.
Spatial Reasoning with Vision-Language Models in Ego-Centric Multi-View Scenes
CV and Pattern Recognition
Helps robots understand 3D space from their own eyes.
Think with 3D: Geometric Imagination Grounded Spatial Reasoning from Limited Views
CV and Pattern Recognition
Lets computers imagine 3D shapes from pictures.