Adrien Bolling

ASCENSION: Autoencoder-Based Latent Space Class Expansion for Time Series Data Augmentation

Tackling the Generative learning trilemma through VAE and GMM-controlled latent space class expansion