Metraipolis › archive › dc_adf080f83e8b

FisherDistill: Diffusion Distribution Distillation via Closed-Form Fisher Critics — arXiv:2611.07604

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.07604
License (as archived): CC0-1.0
Archived: 2026-11-06T14:42:28.920Z
SHA-256 of the archived text: 945bb9ad7d51e4ded4cf3df7aaa8c33d4eb2f4990405dab7129ee308eb167f8a
Stable URL of this page: https://metraipolis.com/documento/dc_adf080f83e8b

How to cite

FisherDistill: Diffusion Distribution Distillation via Closed-Form Fisher Critics — arXiv:2611.07604 — arXiv (en), https://arxiv.org/abs/2611.07604 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_adf080f83e8b (Metraipolis archive dc_adf080f83e8b, sha256 945bb9ad7d51e4ded4cf3df7aaa8c33d4eb2f4990405dab7129ee308eb167f8a)

BibTeX

@misc{dcadf080f83e8b,
  title = {FisherDistill: Diffusion Distribution Distillation via Closed-Form Fisher Critics — arXiv:2611.07604},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.07604},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_adf080f83e8b; sha256 945bb9ad7d51e4ded4cf3df7aaa8c33d4eb2f4990405dab7129ee308eb167f8a},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.07604. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Recent advances in diffusion distillation have enabled high-quality image generation in only one or a few network function evaluations (NFEs). Despite their efficiency, aggressively distilled generators often retain a residual distribution gap from their multi-step teachers, motivating recent methods to introduce adversarial refinement with a learned discriminator. In this work, we ask whether such refinement necessarily requires a learned adversarial discriminator. We introduce \textbf{FisherDistill}, a simple alternative based on Fisher linear discriminant analysis. Using intermediate representations from the frozen teacher, we estimate the first- and second-order statistics of real and generated samples and derive a closed-form Fisher direction that maximally separates their feature distributions relative to within-class variation. This direction provides a lightweight refinement signal that complements the original distillation objective without requiring an additional adversarial discriminator or its optimization loop. Across multiple distillation frameworks, our approach consistently improves generation quality over their respective distilled baselines and substantially closes the gap to adversarially refined counterparts. Code: this https URL

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2611.07604 — 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 2026-11-06T14:18:42.680Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).