Locally Masked Convolution for Autoregressive Models
arXiv:2006.12486
Abstract
High-dimensional generative models have many applications including image compression, multimedia generation, anomaly detection and data completion. State-of-the-art estimators for natural images are autoregressive, decomposing the joint distribution over pixels into a product of conditionals parameterized by a deep neural network, e.g. a convolutional neural network such as the PixelCNN. However, PixelCNNs only model a single decomposition of the joint, and only a single generation order is efficient. For tasks such as image completion, these models are unable to use much of the observed context. To generate data in arbitrary orders, we introduce LMConv: a simple modification to the standard 2D convolution that allows arbitrary masks to be applied to the weights at each location in the image. Using LMConv, we learn an ensemble of distribution estimators that share parameters but differ in generation order, achieving improved performance on whole-image density estimation (2.89 bpd on unconditional CIFAR10), as well as globally coherent image completions. Our code is available at https://ajayjain.github.io/lmconv.
Published at Conference on Uncertainty in AI (UAI) 2020
References in corpus (16)
- PyTorch: An Imperative Style, High-Performance Deep Learning Library
- Auto-Encoding Variational Bayes
- Neural Discrete Representation Learning
- Conditional Image Generation with PixelCNN Decoders
- Pixel Recurrent Neural Networks
- Variational Inference with Normalizing Flows
- Generating Long Sequences with Sparse Transformers
- MADE: Masked Autoencoder for Distribution Estimation
- Deep Generative Stochastic Networks Trainable by Backprop
- Deep AutoRegressive Networks
- RNADE: The real-valued neural autoregressive density-estimator
- A Deep and Tractable Density Estimator
- Iterative Neural Autoregressive Distribution Estimator (NADE-k)
- Neural Autoregressive Distribution Estimation
- Predictive Sampling with Forecasting Autoregressive Models
- Spatial PixelCNN: Generating Images from Patches