Description
Discriminative Regularization is a regularization technique for variational autoencoders that uses representations from discriminative classifiers to augment the VAE objective function (the lower bound) corresponding to a generative model. Specifically, it encourages the model’s reconstructions to be close to the data example in a representation space defined by the hidden layers of highly-discriminative, neural network based classifiers.
Papers Using This Method
Multi-Class Textual-Inversion Secretly Yields a Semantic-Agnostic Classifier2024-10-29Why Fine-Tuning Struggles with Forgetting in Machine Unlearning? Theoretical Insights and a Remedial Approach2024-10-04Epicardium Prompt-guided Real-time Cardiac Ultrasound Frame-to-volume Registration2024-06-20Discriminative Hamiltonian Variational Autoencoder for Accurate Tumor Segmentation in Data-Scarce Regimes2024-06-17Adaptive Discriminative Regularization for Visual Classification2022-03-02Discriminative Regularization for Generative Models2016-02-09