aiqbench analysis
GeoVR addresses the critical limitation of current Multimodal Large Language Models by enabling intrinsic 3D spatial awareness using only 2D video inputs. The framework achieves this by distilling geometric knowledge from pre-trained 3D foundations through a multi-objective strategy that explicitly optimizes for camera pose, depth, scale, and feature alignment. This approach offers a viable pathway to enhance spatial reasoning capabilities in foundation models without relying on scarce large-scale 3D datasets.
Learning Geometric Representations from Videos for Spatial Intelligent Multimodal Large Language Models
arXiv:2606.05833v1 Announce Type: cross Abstract: Multimodal Large Language Models (MLLMs) excel at 2D semantic understanding but lack intrinsic 3D awareness, resulting in representations that fail to maintain geometric and spatial consistency across video frames. Given the scarcity of large-scale 3D data, we present GeoVR, a novel framework that learns geometric representations using purely 2D video sequences. This approach effectively restructures the semantic latent space within MLLMs to unlock spatial intelligence. Rather than employing superficial feature mixing, GeoVR reshapes the internal representations of the MLLM by distilling geometry knowledge from pre-trained 3D foundation models. This is accomplished through a multi-objective learning strategy driven by four complementary geometric targets: (1) estimating inter-frame camera poses to embed varying viewpoint dynamics, (2) regressing dense depth maps to anchor physical distances, (3) predicting a metric scale factor for real-world calibration, and (4) distilling multi-scale 3D features to align the intermediate feature space. Guided by these explicit physical and geometric constraints, the model's internal representations naturally develop strong 3D awareness. Extensive experiments on spatial reasoning benchmarks demonstrate that GeoVR achieves state-of-the-art performance, establishing a new paradigm for endowing foundation models with spatial intelligence.
Abstract:Multimodal Large Language Models (MLLMs) excel at 2D semantic understanding but lack intrinsic 3D awareness, resulting in representations that fail to maintain geometric and spatial consistency across video frames. Given the scarcity of large-scale 3D data, we present GeoVR, a novel framework that learns geometric representations using purely 2D video sequences. This approach effectively restructures the semantic latent space within MLLMs to unlock spatial intelligence. Rather than employing superficial feature mixing, GeoVR reshapes the internal representations of the MLLM by distilling geometry knowledge from pre-trained 3D foundation models. This is accomplished through a multi-objective learning strategy driven by four complementary geometric targets: (1) estimating inter-frame camera poses to embed varying viewpoint dynamics, (2) regressing dense depth maps to anchor physical distances, (3) predicting a metric scale factor for real-world calibration, and (4) distilling multi-scale 3D features to align the intermediate feature space. Guided by these explicit physical and geometric constraints, the model's internal representations naturally develop strong 3D awareness. Extensive experiments on spatial reasoning benchmarks demonstrate that GeoVR achieves state-of-the-art performance, establishing a new paradigm for endowing foundation models with spatial intelligence.
Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
arXiv:2606.05833 [cs.CV]
(or arXiv:2606.05833v1 [cs.CV] for this version)
https://doi.org/10.48550/arXiv.2606.05833 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 282 words
Related coverage
GeoWorld-VLM: Geometry from World Models for Vision-Language Models
arXiv · Jun 12, 2026
NVIDIA Releases Cosmos 3 Edge: A 4B-Parameter Open World Model That Reasons and Generates Robot Actions On-Device
MarkTechPost · Jul 21, 2026
LAYERSCOPE: A Layerwise Characterization of Video and Multimodal Learned Representations
arXiv · Sep 24, 2026
From Symbolic to Geometric: Enabling Spatial Reasoning in Large Language Models
arXiv · Jun 5, 2026