Metraipolis › archive › dc_cf005d9e3aa6

From Static to Dynamic: On-Policy Distillation from Image to Video Diffusion Models — arXiv:2609.34371

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.34371
License (as archived)
CC0-1.0
Archived
2026-10-07T21:02:28.227Z
SHA-256 of the archived text
41b6e9538a752bf3f3164a79541cc31d018506095c088cff0c9586789d2d0060
Stable URL of this page
https://metraipolis.com/documento/dc_cf005d9e3aa6

How to cite

From Static to Dynamic: On-Policy Distillation from Image to Video Diffusion Models — arXiv:2609.34371 — arXiv (en), https://arxiv.org/abs/2609.34371 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_cf005d9e3aa6 (Metraipolis archive dc_cf005d9e3aa6, sha256 41b6e9538a752bf3f3164a79541cc31d018506095c088cff0c9586789d2d0060)
BibTeX
@misc{dccf005d9e3aa6,
  title = {From Static to Dynamic: On-Policy Distillation from Image to Video Diffusion Models — arXiv:2609.34371},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.34371},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_cf005d9e3aa6; sha256 41b6e9538a752bf3f3164a79541cc31d018506095c088cff0c9586789d2d0060},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.34371. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

On-policy distillation (OPD) specializes pretrained video diffusion models through teacher supervision along the student's own generation trajectory. Although large video models are natural teachers, developing specialized video experts can require costly video data and training, while querying them incurs substantially higher latency than querying image experts. More readily available and cheaper to query, image experts offer a cost-effective alternative, particularly for largely temporal-agnostic capabilities such as aesthetics and OCR that admit frame-level supervision. However, heterogeneous image and video latent spaces prevent direct supervision of intermediate student states, while image experts lack cross-frame motion supervision, making temporal consistency vulnerable to frame-level improvements. In this paper, we propose MILD, a Motion-Preserving Image-to-Video Latent Distillation framework that transfers specialized image expertise while preserving pretrained video dynamics. MILD uses a learnable linear connector that aligns student latent states and predicted updates with those of image experts, enabling supervision transfer across heterogeneous latent spaces. We further constrain image-guided corrections around the pretrained student's predictions to preserve video dynamics and incorporate an optical-flow-based motion reward to improve motion quality and temporal consistency. Across specialized image experts and multiple video-student backbones, our method consistently outperforms video-teacher OPD baselines, with further studies demonstrating effective transfer across connector designs and heterogeneous architectures. These results establish image-to-video distillation as an effective route to improving video generation by drawing on the diverse and evolving capabilities of the image-generation ecosystem.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.34371 — 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 2026-10-07T20:29:02.965Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).