Abstract

PixelCNN achieves state-of-the-art results in density estimation for natural images. Although training is fast, inference is costly, requiring on network evaluation per pixel; O(N) for N pixels. This can be sped up by caching activations, but still involves generating each pixel sequentially. In this work, we propose a parallelized PixelCNN that allows more efficient inference by modeling certain pixel groups as conditionally independent. Our new PixelCNN model achieves competitive density estimation and orders of magnitude speedup -O(log N) sampling instead of O(N) enabling the practical generation of 512 x 512 images. We evaluate the model on class-conditional image generation, text-toimage synthesis, and action-conditional video generation, showing that our model achieves the best results among non-pixel-autoregressive density models that allow efficient sampling.