{"ID":2837897,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2511.18281","arxiv_id":"2511.18281","title":"Uni-DAD: Unified Distillation and Adaptation of Diffusion Models for Few-step Few-shot Image Generation","abstract":"Diffusion models (DMs) produce high-quality images, yet their sampling remains costly when adapted to new domains. Distilled DMs are faster but typically remain confined within their teacher's domain. Thus, fast and high-quality generation for novel domains relies on two-stage pipelines: Adapt-then-Distill or Distill-then-Adapt. However, both add design complexity and often degrade quality or diversity. We introduce Uni-DAD, a single-stage pipeline that unifies DM distillation and adaptation. It couples two training signals: (i) a dual-domain distribution-matching distillation (DMD) objective that guides the student toward the distributions of the source teacher and a target teacher, and (ii) a multi-head generative adversarial network (GAN) loss that encourages target realism across multiple feature scales. The source domain distillation preserves diverse source knowledge, while the multi-head GAN stabilizes training and reduces overfitting, especially in few-shot regimes. The inclusion of a target teacher facilitates adaptation to more structurally distant domains. We evaluate Uni-DAD on two comprehensive benchmarks for few-shot image generation (FSIG) and subject-driven personalization (SDP) using diffusion backbones. It delivers better or comparable quality to state-of-the-art (SoTA) adaptation methods even with less than 4 sampling steps, and often surpasses two-stage pipelines in quality and diversity. Code: https://github.com/yaramohamadi/uni-DAD.","short_abstract":"Diffusion models (DMs) produce high-quality images, yet their sampling remains costly when adapted to new domains. Distilled DMs are faster but typically remain confined within their teacher's domain. Thus, fast and high-quality generation for novel domains relies on two-stage pipelines: Adapt-then-Distill or Distill-t...","url_abs":"https://arxiv.org/abs/2511.18281","url_pdf":"https://arxiv.org/pdf/2511.18281v3","authors":"[\"Yara Bahram\",\"Mélodie Desbos\",\"Mohammadhadi Shateri\",\"Eric Granger\"]","published":"2025-11-23T04:22:42Z","proceeding":"cs.CV","tasks":"[\"cs.CV\",\"cs.AI\"]","methods":"[\"Diffusion Model\",\"Generative Adversarial Network\"]","has_code":false,"code_links":[{"ID":606721,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2837897,"paper_url":"https://arxiv.org/abs/2511.18281","paper_title":"Uni-DAD: Unified Distillation and Adaptation of Diffusion Models for Few-step Few-shot Image Generation","repo_url":"https://github.com/yaramohamadi/uni-DAD","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}