Сравнение на методи

	Генеративни състезателни мрежи с трансферно обучение (Transfer Learning GAN)	Трансферно обучение с дифузионен модел
Област	Дълбоко обучение	Дълбоко обучение
Семейство	Machine learning	Machine learning
Година на възникване	2014–2018	2020–2023
Създател	Goodfellow, I. et al. (GAN); Wang & Ramanan (transfer to GAN)	Ho et al. (DDPM); transfer application popularized by Rombach et al. (Stable Diffusion) and Ruiz et al. (DreamBooth), 2020–2023
Тип	Generative model with transferred weights	Generative model with transfer learning
Основополагащ източник	Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A. & Bengio, Y. (2014). Generative Adversarial Nets. Advances in Neural Information Processing Systems (NeurIPS), 27, 2672–2680.	Ho, J., Jain, A., & Abbeel, P. (2020). Denoising Diffusion Probabilistic Models. Advances in Neural Information Processing Systems (NeurIPS), 33, 6840–6851.
Други названия	TL-GAN, pretrained GAN, GAN fine-tuning, domain-adaptive GAN	diffusion model fine-tuning, pre-trained diffusion transfer, TL-DM, domain-adapted diffusion model
Свързани	6	5
Резюме	Transfer Learning GAN initialises a Generative Adversarial Network — or both its generator and discriminator — from weights pretrained on a large source dataset, then fine-tunes the network on a smaller target dataset. This approach allows high-quality generative modelling even when target-domain data are scarce, by reusing low- and mid-level feature representations learned at scale.	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.