Metraipolis › archive › dc_bb1508848279

AR-Edit: Training-Free Autoregressive Video Editing without Inversion — arXiv:2611.09795

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.09795
License (as archived)
CC0-1.0
Archived
2026-11-09T22:45:29.201Z
SHA-256 of the archived text
131e8bcc5927d6925b5eb09c82131e2b85d8ee4004d53e9ec049132197030d23
Stable URL of this page
https://metraipolis.com/documento/dc_bb1508848279

How to cite

AR-Edit: Training-Free Autoregressive Video Editing without Inversion — arXiv:2611.09795 — arXiv (en), https://arxiv.org/abs/2611.09795 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_bb1508848279 (Metraipolis archive dc_bb1508848279, sha256 131e8bcc5927d6925b5eb09c82131e2b85d8ee4004d53e9ec049132197030d23)
BibTeX
@misc{dcbb1508848279,
  title = {AR-Edit: Training-Free Autoregressive Video Editing without Inversion — arXiv:2611.09795},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.09795},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_bb1508848279; sha256 131e8bcc5927d6925b5eb09c82131e2b85d8ee4004d53e9ec049132197030d23},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.09795. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Editing a video stream in real time, without access to future frames, is essential for interactive applications. Existing training-free video editing methods, however, assume full offline access to the entire sequence, while streaming approaches either rely on additional training or are limited to specific editing techniques. We show that this limitation can be removed by analyzing the self-attention dynamics of DMD-distilled autoregressive (AR) video diffusion models. We uncover a key asymmetry: queries and keys remain stable and aligned with their source counterparts across denoising timesteps, while values diverge and encode the evolving content. Inherited from the non-distilled bidirectional teacher, this asymmetry is strongest at high noise, where few-step streaming models operate. The resulting mechanism is simple: a single forward pass suffices to capture source identity in the value features, which can be reused throughout generation. Starting from pure Gaussian noise and injecting only these source values reconstructs the input with fidelity matching or surpassing inversion-based methods, effectively eliminating inversion. Based on this insight, we introduce AR-Edit, a training-free, inversion-free method that caches source values once and selectively injects them during denoising. Across multiple AR backbones, AR-Edit enables real-time streaming editing with superior source preservation and tonal fidelity at minimal overhead.

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/2611.09795 — 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 2026-11-09T22:37:47.234Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).