SpatialBoost: Enhancing Visual Representation through Language-Guided Reasoning

TL;DR

SpatialBoost enhances 2D vision models by integrating 3D spatial knowledge via language, significantly improving their performance on perception benchmarks through a hierarchical reasoning process.

Contribution

It introduces a scalable framework that injects 3D spatial understanding into pre-trained vision encoders using language and multi-turn reasoning.

Findings

Improves DINOv3 mIoU on ADE20K from 55.9 to 59.7.

Achieves state-of-the-art performance with 3.8% gain.

Demonstrates effectiveness across various vision benchmarks.

Abstract

Despite the remarkable success of large-scale pre-trained image representation models (i.e., vision encoders) across various vision tasks, they are predominantly trained on 2D image data and therefore often fail to capture 3D spatial relationships between objects and backgrounds in the real world, constraining their effectiveness in many downstream applications. To address this, we propose SpatialBoost, a scalable framework that enhances the spatial awareness of existing pre-trained vision encoders by injecting 3D spatial knowledge expressed in linguistic descriptions. The core idea involves converting dense 3D spatial information from 2D images into linguistic expressions, which is then used to inject such spatial knowledge into vision encoders through a Large Language Model (LLM). To this end, we adopt a multi-turn Chain-of-Thought (CoT) reasoning process that progressively incorporates dense spatial knowledge and builds hierarchical spatial understanding. To validate effectiveness, we adapt SpatialBoost to state-of-the-art vision encoders such as DINOv3, and evaluate its performance gains on a wide range of benchmarks requiring both 3D perception and general vision abilities. For instance, SpatialBoost improves DINOv3 performance from 55.9 to 59.7 mIoU on ADE20K, achieving state-of-the-art performance with 3.8% gain over the pre-trained DINOv3.