World2VLM: Distilling World Model Imagination into VLMs for Dynamic Spatial Reasoning
Vision-language models (VLMs) have shown strong performance on static visual understanding, yet they still struggle with dynamic spatial reasoning that requires imagining how scenes evolve under egocentric motion. Recent efforts address this limitation either by scaling spatial supervision with synthetic data or by coupling VLMs with world models at inference time. However, the former often lacks explicit modeling of motion-conditioned state transitions, while the latter incurs substantial computational overhead. In this work, we propose World2VLM, a training framework that distills spatial imagination from a generative world model into a vision-language model. Given an initial observatio...
| Authors | Wanyue Zhang, Wenxiang Wu, Wang Xu et al. |
|---|---|
| Primary category | cs.CV |
| Published | 29 April 2026 |
| Last updated at source | 29 April 2026 |
Models and datasets referencing this paper
- WanyueZhang/World2VLMDataset
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.