Comparer des méthodes

	Apprentissage par transfert avec autoencodeur variationnel	Variational Autoencoder affiné
Domaine	Apprentissage profond	Apprentissage profond
Famille	Machine learning	Machine learning
Année d'origine	2014 (VAE); 2010 (transfer learning survey)	2014 (VAE); fine-tuning practice from 2015 onward
Auteur d'origine	Kingma, D. P. & Welling, M. (VAE); transfer learning framework from Pan & Yang	Kingma, D. P. & Welling, M. (VAE); fine-tuning strategy from transfer learning literature
Type	Generative model with transferred encoder/decoder	Generative model with fine-tuning
Source fondatrice	Kingma, D. P., & Welling, M. (2014). Auto-Encoding Variational Bayes. International Conference on Learning Representations (ICLR 2014).	Kingma, D. P., & Welling, M. (2014). Auto-Encoding Variational Bayes. In Proceedings of the 2nd International Conference on Learning Representations (ICLR 2014).
Alias	TL-VAE, pretrained VAE, VAE transfer learning, fine-tuned variational autoencoder	fine-tuned VAE, domain-adapted VAE, transfer-learned VAE, adapted variational autoencoder
Apparentées	6	6
Résumé	Transfer Learning with a Variational Autoencoder (TL-VAE) reuses an encoder and/or decoder pre-trained on a large source dataset and adapts it to a smaller target domain. By inheriting a rich probabilistic latent space rather than starting from random weights, TL-VAE dramatically reduces the amount of target-domain data needed for high-quality generation or representation learning.	A Fine-Tuned Variational Autoencoder begins with a VAE pre-trained on a large source dataset and then continues training on a smaller target-domain dataset. This approach adapts the learned latent representation and generative capacity to new data, preserving general structure while specializing to the target distribution — yielding better results than training from scratch when labeled or large target data is scarce.