NVAE, or Nouveau VAE, is deep, hierarchical variational autoencoder. It can be trained with the original VAE objective, unlike alternatives such as VQ-VAE-2. NVAE’s design focuses on tackling two main challenges: (i) designing expressive neural networks specifically for VAEs, and (ii) scaling up the training to a large number of hierarchical groups and image sizes while maintaining training stability.
To tackle long-range correlations in the data, the model employs hierarchical multi-scale modelling. The generative model starts from a small spatially arranged latent variables as $\mathbf{z}_{1}$ and samples from the hierarchy group-by-group while gradually doubling the spatial dimensions. This multi-scale approach enables NVAE to capture global long-range correlations at the top of the hierarchy and local fine-grained dependencies at the lower groups.
Additional design choices include the use of residual cells for the generative models and the encoder, which employ a number of tricks and modules to achieve good performance, and the use of residual normal distributions to smooth optimization. See the components section for more details.
Source: NVAE: A Deep Hierarchical Variational AutoencoderPaper | Code | Results | Date | Stars |
---|
Task | Papers | Share |
---|---|---|
Image Generation | 2 | 33.33% |
Adversarial Attack | 1 | 16.67% |
Anomaly Detection | 1 | 16.67% |
Time Series Analysis | 1 | 16.67% |
Time Series Anomaly Detection | 1 | 16.67% |