2023/01/31 by Omead Pooladzandi, Pasha Khosravi, Pooladzandi, Omead +5
Computer Science · #Generative Adversarial Networks and Image Synthesis #Neural Networks and Applications #Music and Audio Processing
paper · pdf · doi:10.48550/arxiv.2302.00138
Generative models have the ability to synthesize data points drawn from the data distribution, however, not all generated samples are high quality. In this paper, we propose using a combination of coresets selection methods and ``entropic regularization'' to select the highest fidelity samples. We leverage an Energy-Based Model which resembles a variational auto-encoder with an inference and generator model for which the latent prior is complexified by an energy-based model. In a semi-supervised learning scenario, we show that augmenting the labeled data-set, by adding our selected subset of samples, leads to better accuracy improvement rather than using all the synthetic samples.