Corpus ID: 17306137

PixelVAE: A Latent Variable Model for Natural Images

@article{Gulrajani2017PixelVAEAL,
  title={PixelVAE: A Latent Variable Model for Natural Images},
  author={Ishaan Gulrajani and K. Kumar and F. Ahmed and Adrien Ali Ta{\"i}ga and Francesco Visin and D. V{\'a}zquez and A. Courville},
  journal={ArXiv},
  year={2017},
  volume={abs/1611.05013}
}
  • Ishaan Gulrajani, K. Kumar, +4 authors A. Courville
  • Published 2017
  • Mathematics, Computer Science
  • ArXiv
  • Natural image modeling is a landmark challenge of unsupervised learning. [...] Key Method Our model requires very few expensive autoregressive layers compared to PixelCNN and learns latent codes that are more compressed than a standard VAE while still capturing most non-trivial structure. Finally, we extend our model to a hierarchy of latent variables at different scales. Our model achieves state-of-the-art performance on binarized MNIST, competitive performance on 64 × 64 ImageNet, and high-quality samples on…Expand Abstract

    Figures, Tables, and Topics from this paper.

    Paper Mentions

    Neural Discrete Representation Learning
    • 422
    • Open Access
    VAE with a VampPrior
    • 244
    • Highly Influenced
    • Open Access
    Joint Autoregressive and Hierarchical Priors for Learned Image Compression
    • 145
    • Highly Influenced
    • Open Access
    A Hierarchical Latent Vector Model for Learning Long-Term Structure in Music
    • 138
    • Open Access
    InfoVAE: Information Maximizing Variational Autoencoders
    • 211
    • Highly Influenced
    • Open Access
    PixelGAN Autoencoders
    • 53
    • Open Access
    Parallel Multiscale Autoregressive Density Estimation
    • 99
    • Open Access
    PixelSNAIL: An Improved Autoregressive Generative Model
    • 69
    • Open Access
    Recent Advances in Autoencoder-Based Representation Learning
    • 100
    • Highly Influenced
    • Open Access

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 32 REFERENCES
    Pixel Recurrent Neural Networks
    • 1,218
    • Open Access
    Generative Adversarial Nets
    • 17,548
    • Open Access
    Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
    • 6,478
    • Open Access
    Auto-Encoding Variational Bayes
    • 9,051
    • Open Access
    Variational Lossy Autoencoder
    • 352
    • Open Access
    Adversarial Feature Learning
    • 889
    • Open Access
    Importance Weighted Autoencoders
    • 628
    • Open Access
    Ladder Variational Autoencoders
    • 396
    • Open Access
    Density estimation using Real NVP
    • 913
    • Open Access