Generating High Fidelity Synthetic Data via Coreset selection and Entropic Regularization
This work addresses data quality issues in semi-supervised learning, but it is incremental as it builds on existing generative and selection methods.
The paper tackles the problem of low-quality synthetic data from generative models by selecting high-fidelity samples using coresets and entropic regularization, showing that augmenting labeled datasets with these selected samples improves accuracy more than using all synthetic samples.
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.