方法对比

	迁移学习GAN	扩散模型迁移学习
领域	深度学习	深度学习
方法族	Machine learning	Machine learning
起源年份	2014–2018	2020–2023
提出者	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
类型	Generative model with transferred weights	Generative model with transfer learning
开创性文献	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.
别名	TL-GAN, pretrained GAN, GAN fine-tuning, domain-adaptive GAN	diffusion model fine-tuning, pre-trained diffusion transfer, TL-DM, domain-adapted diffusion model
相关	6	5
摘要	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.