Metraipolis › archive › dc_13fe038ee808

Decoupled Drift-Diffusion World Models for Consistent Future Frame Prediction — arXiv:2610.19212

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.19212
License (as archived)
CC0-1.0
Archived
2026-10-15T21:55:43.558Z
SHA-256 of the archived text
29567bd5439c6651f377027c0b9e66a5b137e370e869913b49ee35ad1685b28b
Stable URL of this page
https://metraipolis.com/documento/dc_13fe038ee808

How to cite

Decoupled Drift-Diffusion World Models for Consistent Future Frame Prediction — arXiv:2610.19212 — arXiv (en), https://arxiv.org/abs/2610.19212 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_13fe038ee808 (Metraipolis archive dc_13fe038ee808, sha256 29567bd5439c6651f377027c0b9e66a5b137e370e869913b49ee35ad1685b28b)
BibTeX
@misc{dc13fe038ee808,
  title = {Decoupled Drift-Diffusion World Models for Consistent Future Frame Prediction — arXiv:2610.19212},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.19212},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_13fe038ee808; sha256 29567bd5439c6651f377027c0b9e66a5b137e370e869913b49ee35ad1685b28b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.19212. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

World models of visual information predict future image frames conditioned on given action sequences. This predicted scene evolution is driven by deterministic changes such as controlled motion on a robot platform, as well as stochastic changes such as rendering of manipulated objects or previously occluded regions. Our key insight is that the spatial location of these stochastic changes is itself learnable. We leverage this insight by jointly regressing both the deterministic next-state prediction and the spatial uncertainty map, then using the uncertainty map to guide a conditional diffusion process. We propose Decoupled Drift-Diffusion World Model (D3WM), which uses a transformer trained with a Gaussian negative-log-likelihood objective to produce both the deterministic next-state prediction and a spatial map of heteroscedastic aleatoric uncertainty. The learned uncertainty map then provides a structured spatial gate for a conditional diffusion model: deterministic regions are anchored to the predicted state, while uncertain regions, such as disocclusions or dynamic textures, are sampled by the diffusion process. We frame this decomposition under a variance-gated stochastic differential equation that separates deterministic drift from stochastic diffusion. D3WM shows improved results over competitive baselines in two distinct types of action spaces: robotic tabletop manipulation and camera motion.

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2610.19212 — 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 2026-10-15T21:38:29.446Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).