Metraipolis › archive › dc_02800b2da208
Spatially Aware World Action Model via Geometric Latent Diffusion — arXiv:2609.02531
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.02531
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T05:50:08.484Z
- SHA-256 of the archived text
807d53470664d399a40162902efd62c6140797bf80a8c6558190387e30d710e5- Stable URL of this page
https://metraipolis.com/documento/dc_02800b2da208
How to cite
Spatially Aware World Action Model via Geometric Latent Diffusion — arXiv:2609.02531 — arXiv (en), https://arxiv.org/abs/2609.02531 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_02800b2da208 (Metraipolis archive dc_02800b2da208, sha256 807d53470664d399a40162902efd62c6140797bf80a8c6558190387e30d710e5)
BibTeX
@misc{dc02800b2da208,
title = {Spatially Aware World Action Model via Geometric Latent Diffusion — arXiv:2609.02531},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.02531},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_02800b2da208; sha256 807d53470664d399a40162902efd62c6140797bf80a8c6558190387e30d710e5},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.02531. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
World Action Models (WAMs) leverage the capabilities of large-scale pretrained video diffusion models to jointly predict future observations and actions, inheriting rich visual and physical priors from internet-scale video. This has made them a promising paradigm for robot policy learning, yet the prevailing models operate exclusively on RGB observations and do not leverage 3D information. To bridge this gap, we introduce a Spatially Aware World Action Model (SA-WAM), which repurposes a pretrained video model for joint action, RGB, and depth prediction, enabling 3D-aware world modeling and action prediction within a single diffusion backbone. We use a nonlinear encoding that maps the unbounded depth signal into the bounded input domain expected by the frozen VAE tokenizer. This allows us to reuse the tokenizer without 3D-specific fine-tuning, incorporating geometric information without sacrificing the pretrained priors. SA-WAM achieves state-of-the-art results on the RoboCasa and LIBERO-Plus benchmarks, while simultaneously improving future-state predictions. Furthermore, SA-WAM outperforms strong baselines in real-world evaluation using a UR5 robotic arm, with strong gains in randomized environments. We analyze the correlation between world model prediction quality and rollout success, providing insights into WAM performance and avenues for its improvement. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-03 in cs.CV). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Robotics (cs.RO). Pagina del lavoro: https://arxiv.org/abs/2609.02531 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-17T05:25:58.131Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-03?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).