Metraipolis › archive › dc_35fab96be306

LongTake: Learning to Sustain Dynamics in Long-Horizon Video Generation — arXiv:2609.38562

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.38562
License (as archived)
CC0-1.0
Archived
2026-10-07T14:45:22.247Z
SHA-256 of the archived text
0c6e68c320cc8e0648efd3e4312ded5d0546d04f7a61605c039ae84bbed872b5
Stable URL of this page
https://metraipolis.com/documento/dc_35fab96be306

How to cite

LongTake: Learning to Sustain Dynamics in Long-Horizon Video Generation — arXiv:2609.38562 — arXiv (en), https://arxiv.org/abs/2609.38562 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_35fab96be306 (Metraipolis archive dc_35fab96be306, sha256 0c6e68c320cc8e0648efd3e4312ded5d0546d04f7a61605c039ae84bbed872b5)
BibTeX
@misc{dc35fab96be306,
  title = {LongTake: Learning to Sustain Dynamics in Long-Horizon Video Generation — arXiv:2609.38562},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.38562},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_35fab96be306; sha256 0c6e68c320cc8e0648efd3e4312ded5d0546d04f7a61605c039ae84bbed872b5},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.38562. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

World models, game simulators, and long-take video creation require coherent scene evolution and sustained dynamics over extended durations. Autoregressive (AR) video diffusion provides a natural framework for long-horizon generation, yet extended rollouts often become near-static or lose visual quality. We hypothesize that these failures reflect the limited guidance provided by short-video supervision on how ongoing scene dynamics develops over longer durations. This motivates us to introduce LongTake, a two-stage training pipeline built around Long-Horizon Teacher Forcing (TF) on curated real long videos. Long-Horizon TF trains the AR model to predict later frames conditioned on long ground-truth video prefixes, extending direct supervision beyond the short training horizon. This supervision is designed to help the model sustain dynamics and preserve visual quality during long-horizon generation. Our central finding is that this training stage strengthens direct initialization for distribution matching distillation (DMD) under student self-rollout, without the intermediate few-step distillation stage used in standard pipelines. Under the same five-second DMD training setup, our initialization yields substantially higher dynamic degree than short horizon TF initialization on 30-second rollouts at comparable aesthetic quality, and surpasses the evaluated baselines in both measures. Hybrid DMD further reuses this teacher to extend supervision to later frames of the self-rollout while retaining bidirectional joint supervision over the initial window. On long-horizon self-rollouts, LongTake lies on the Pareto front of dynamic degree and aesthetic quality, and Hybrid DMD attains the highest dynamic degree among evaluated methods at both 30s and 60s.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-10-01 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.38562 — 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 2026-10-07T14:19:00.575Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).