Σύγκριση μεθόδων
Εξετάστε τις επιλεγμένες μεθόδους δίπλα-δίπλα· οι γραμμές που διαφέρουν επισημαίνονται.
| Μοντέλο Διάχυσης Προσαρμοσμένο στον Τομέα× | Μοντέλα Διάχυσης και Μεταφορά Μάθησης× | |
|---|---|---|
| Πεδίο | Βαθιά Μάθηση | Βαθιά Μάθηση |
| Οικογένεια | Machine learning | Machine learning |
| Έτος προέλευσης≠ | 2022–2023 | 2020–2023 |
| Δημιουργός≠ | Ho et al. (DDPM); domain-adaptation variants popularized by Gal et al. and Ruiz et al. (2022–2023) | Ho et al. (DDPM); transfer application popularized by Rombach et al. (Stable Diffusion) and Ruiz et al. (DreamBooth), 2020–2023 |
| Τύπος≠ | Generative model with domain adaptation | Generative model with transfer learning |
| Θεμελιώδης πηγή≠ | Ho, J., Jain, A., & Abbeel, P. (2020). Denoising Diffusion Probabilistic Models. Advances in Neural Information Processing Systems, 33, 6840–6851. link ↗ | Ho, J., Jain, A., & Abbeel, P. (2020). Denoising Diffusion Probabilistic Models. Advances in Neural Information Processing Systems (NeurIPS), 33, 6840–6851. link ↗ |
| Εναλλακτικές ονομασίες | DA-diffusion model, domain-adapted diffusion model, domain-adaptive DDPM, cross-domain diffusion model | diffusion model fine-tuning, pre-trained diffusion transfer, TL-DM, domain-adapted diffusion model |
| Συναφείς≠ | 6 | 5 |
| Σύνοψη≠ | A domain-adaptive diffusion model is a denoising diffusion probabilistic model (DDPM) that is pre-trained on large general datasets and then adapted — through fine-tuning, textual inversion, or LoRA — to generate high-quality outputs in a specific target domain. It combines the powerful generative capacity of diffusion models with domain adaptation techniques, enabling high-fidelity synthesis in specialized areas such as medical imaging, satellite imagery, or domain-specific art styles with limited target-domain data. | Transfer Learning with Diffusion Models adapts a large pre-trained diffusion model — such as Stable Diffusion or DALL-E 2 — to a new target domain or task by continuing training on a smaller domain-specific dataset. Rather than learning the full generative process from scratch, practitioners leverage knowledge already encoded in millions of training steps to achieve high-quality domain-adapted generation with modest data and compute. |
| ScholarGateΣύνολο δεδομένων ↗ |
|
|