Spatial Reasoning with Denoising Models
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866916789616640000 |
|---|---|
| author | Wewer, Christopher Pogodzinski, Bart Schiele, Bernt Lenssen, Jan Eric |
| author_facet | Wewer, Christopher Pogodzinski, Bart Schiele, Bernt Lenssen, Jan Eric |
| contents | We introduce Spatial Reasoning Models (SRMs), a framework to perform reasoning over sets of continuous variables via denoising generative models. SRMs infer continuous representations on a set of unobserved variables, given observations on observed variables. Current generative models on spatial domains, such as diffusion and flow matching models, often collapse to hallucination in case of complex distributions. To measure this, we introduce a set of benchmark tasks that test the quality of complex reasoning in generative models and can quantify hallucination. The SRM framework allows to report key findings about importance of sequentialization in generation, the associated order, as well as the sampling strategies during training. It demonstrates, for the first time, that order of generation can successfully be predicted by the denoising network itself. Using these findings, we can increase the accuracy of specific reasoning tasks from <1% to >50%. Our project website provides additional videos, code, and the benchmark datasets: https://geometric-rl.mpi-inf.mpg.de/srm |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_21075 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Spatial Reasoning with Denoising Models Wewer, Christopher Pogodzinski, Bart Schiele, Bernt Lenssen, Jan Eric Computer Vision and Pattern Recognition Machine Learning We introduce Spatial Reasoning Models (SRMs), a framework to perform reasoning over sets of continuous variables via denoising generative models. SRMs infer continuous representations on a set of unobserved variables, given observations on observed variables. Current generative models on spatial domains, such as diffusion and flow matching models, often collapse to hallucination in case of complex distributions. To measure this, we introduce a set of benchmark tasks that test the quality of complex reasoning in generative models and can quantify hallucination. The SRM framework allows to report key findings about importance of sequentialization in generation, the associated order, as well as the sampling strategies during training. It demonstrates, for the first time, that order of generation can successfully be predicted by the denoising network itself. Using these findings, we can increase the accuracy of specific reasoning tasks from <1% to >50%. Our project website provides additional videos, code, and the benchmark datasets: https://geometric-rl.mpi-inf.mpg.de/srm |
| title | Spatial Reasoning with Denoising Models |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2502.21075 |