方法对比

	自监督扩散模型	生成对抗网络	变分自编码器
领域	深度学习	深度学习	深度学习
方法族	Machine learning	Machine learning	Machine learning
起源年份	2020–2022	2014	2014
提出者	Ho, J. et al.; extended by Chen, T. et al. and subsequent self-supervised diffusion works	Goodfellow, I. et al.	Kingma, D. P. & Welling, M.
类型	Generative model with self-supervised representation objective	Generative deep learning (adversarial two-network game)	Deep generative latent-variable model (encoder–decoder)
开创性文献	Ho, J., Jain, A., & Abbeel, P. (2020). Denoising Diffusion Probabilistic Models. Advances in Neural Information Processing Systems (NeurIPS), 33, 6840–6851.	Goodfellow, I. et al. (2014). Generative Adversarial Nets. NeurIPS.	Kingma, D. P. & Welling, M. (2014). Auto-Encoding Variational Bayes. International Conference on Learning Representations (ICLR).
别名	SSDM, self-supervised score-based model, diffusion-based self-supervised learning, denoising diffusion with self-supervised pretraining	Üretici Çekişmeli Ağ (GAN), GAN, generative adversarial nets, adversarial network	Değişkensel Otokodlayıcı (VAE), VAE, auto-encoding variational Bayes, deep latent variable model
相关	2	4	5
摘要	A self-supervised diffusion model couples the iterative noise-and-denoise generative process of denoising diffusion probabilistic models with a self-supervised representation learning objective — such as contrastive or masked prediction loss — so that the model simultaneously learns to generate realistic data and to produce semantically meaningful representations without any labeled examples.	A Generative Adversarial Network (GAN), introduced by Ian Goodfellow and colleagues in 2014, produces realistic synthetic data through the competition of two neural networks — a generator and a discriminator. It is widely used for image synthesis, data augmentation, and distribution estimation.	The Variational Autoencoder (VAE) is a deep generative latent-variable model, introduced by Diederik Kingma and Max Welling in 2014, that encodes data as a probability distribution in a latent space and samples from that distribution to generate new examples. It is used for data generation, anomaly detection, and feature learning.