Jan Egger

Training β-VAE by Aggregating a Learned Gaussian Posterior with a Decoupled Decoder

2022 · arXiv (Cornell University)

The reconstruction loss and the Kullback-Leibler divergence (KLD) loss in a variational autoencoder (VAE) often play antagonistic roles, and tuning the weight of the KLD loss in $β$-VAE to achieve a balance between the two …