Abstract | ||
---|---|---|
Representation learning seeks to expose certain aspects of observed data in a learned representation that's amenable to downstream tasks like classification. For instance, a good representation for 2D images might be one that describes only global structure and discards information about detailed texture. In this paper, we present a simple but principled method to learn such global representations by combining Variational Autoencoder (VAE) with neural autoregressive models such as RNN, MADE and PixelRNN/CNN. Our proposed VAE model allows us to have control over what the global latent code can learn and , by designing the architecture accordingly, we can force the global latent code to discard irrelevant information such as texture in 2D images, and hence the VAE only "autoencodes" data in a lossy fashion. In addition, by leveraging autoregressive models as both prior distribution $p(z)$ and decoding distribution $p(x|z)$, we can greatly improve generative modeling performance of VAEs, achieving new state-of-the-art results on MNIST, OMNIGLOT and Caltech-101 Silhouettes density estimation tasks. |
Year | Venue | DocType |
---|---|---|
2017 | ICLR | Conference |
Volume | Citations | PageRank |
abs/1611.02731 | 0 | 0.34 |
References | Authors | |
0 | 8 |
Name | Order | Citations | PageRank |
---|---|---|---|
Xi Chen | 1 | 1649 | 54.94 |
Diederik P. Kingma | 2 | 8013 | 263.16 |
Tim Salimans | 3 | 1259 | 47.13 |
Yan Duan | 4 | 775 | 27.97 |
Prafulla Dhariwal | 5 | 83 | 2.88 |
John Schulman | 6 | 1806 | 66.95 |
Ilya Sutskever | 7 | 25814 | 1120.24 |
Pieter Abbeel | 8 | 6363 | 376.48 |