Metraipolis › archive › dc_83d387d8a482

Continuous Adversarial MeanFlow Transfer — arXiv:2608.19540

Source: arXiv (en)
Original: https://arxiv.org/abs/2608.19540
License (as archived): CC0-1.0
Archived: 2026-10-18T11:48:09.689Z
SHA-256 of the archived text: b51735069bc4c95b3e2932ba9fc22e481815414316faf0994143047cb8173784
Stable URL of this page: https://metraipolis.com/documento/dc_83d387d8a482

How to cite

Continuous Adversarial MeanFlow Transfer — arXiv:2608.19540 — arXiv (en), https://arxiv.org/abs/2608.19540 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_83d387d8a482 (Metraipolis archive dc_83d387d8a482, sha256 b51735069bc4c95b3e2932ba9fc22e481815414316faf0994143047cb8173784)

BibTeX

@misc{dc83d387d8a482,
  title = {Continuous Adversarial MeanFlow Transfer — arXiv:2608.19540},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.19540},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_83d387d8a482; sha256 b51735069bc4c95b3e2932ba9fc22e481815414316faf0994143047cb8173784},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.19540. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Training fast generators on new domains with limited data remains challenging for two reasons. First, adapting a pretrained diffusion or flow model to a new domain leaves its costly multi-step sampling unaddressed, and existing acceleration methods are tied to the source parameterization--$\epsilon$, $x$, $v$, or $u$--leaving heterogeneous pretrained models with no common acceleration target. Second, while adversarial refinement is proven effective for few-step quality, it is formulated only for instantaneous-velocity flows, not for the finite-interval average velocities that MeanFlow (MF) models predict. We address both problems. We propose MeanFlow-Transfer, which maps heterogeneous source outputs into a shared velocity representation, uses it to initialize an MF generator from the source weights, and optimizes an MF objective on the target domain. This unifies adaptation and acceleration in a single training loop across a broad range of pretrained models. We then introduce Continuous Adversarial MeanFlow, a post-training stage that extends continuous adversarial flow models from instantaneous velocities to MF's finite-interval average velocities. CAMF contrasts changes in a learned potential between real and predicted interval endpoints, recovering fine detail that MF regression averages away, and reduces to the instantaneous criterion in the vanishing-interval limit. Adapting four ImageNet-based source models--DiT ($\epsilon$), SiT ($v$), JiT ($x$), iMF ($u$)--to five target domains, MF-T with CAMF matches or exceeds the fine-tuned teacher in FID and FDD at up to $125\times$ fewer Neural Function Evaluations (NFEs), while CAMF improves MF-T's few-step FID by $29\%$ on average.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-21 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG) ; Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2608.19540 — 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 2026-10-18T11:47:43.302Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).