SpatialLadder: Progressive Training for Spatial Reasoning in Vision-Language Models
About
Spatial reasoning remains a fundamental challenge for Vision-Language Models (VLMs), with current approaches struggling to achieve robust performance despite recent advances. We identify that this limitation stems from a critical gap: existing methods attempt to learn spatial reasoning directly without establishing the hierarchical foundations of perception and understanding. To address this challenge, we present a comprehensive methodology for building spatial intelligence progressively. We introduce SpatialLadder-26k, a multimodal dataset containing 26,610 samples spanning object localization, single image, multi-view, and video spatial reasoning tasks, constructed through a standardized pipeline that ensures systematic coverage across modalities. Building on this dataset, we design a three-stage progressive training framework that (1) establishes spatial perception through object localization, (2) develops spatial understanding through multi-dimensional spatial tasks, and (3) strengthens complex reasoning via reinforcement learning with verifiable rewards. This approach yields SpatialLadder, a 3B-parameter model that achieves state-of-the-art performance on spatial reasoning benchmarks, with 23.4% average improvement over the base model, surpassing GPT-4o by 20.8% and Gemini-2.0-Flash by 10.1%. Notably, SpatialLadder maintains strong generalization with 7.2% improvement on out-of-domain benchmarks, demonstrating that progressive training from perception to reasoning is essential for robust spatial intelligence.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Spatial Reasoning | CV-Bench | Accuracy73.7 | 46 | |
| Spatial Reasoning | MindCube | Accuracy43.4 | 37 | |
| Multimodal Spatial Intelligence | EASI (In-Domain) | Average Score39.1 | 32 | |
| Spatial Reasoning | MindCube tiny (test) | Rot. Accuracy30.5 | 30 | |
| Spatial Reasoning | MMSI-Bench (test) | PR Score30.3 | 29 | |
| Spatial Reasoning | Viewspatial | Accuracy44.2 | 28 | |
| Spatial Reasoning | VSI-Bench | Accuracy44.8 | 24 | |
| Spatial Reasoning | MMSI-Bench | Accuracy27.4 | 24 | |
| Spatial Reasoning | SITE | Accuracy27.9 | 24 | |
| Spatial Reasoning | CV-Bench-3D | Accuracy74.9 | 21 |