Prof. Scott Reed, gave a lecture titled "Recent Advances in Autoregressive Density Estimation with Deep Neural Networks" at AMSS on 10, November 2017.
Autoregressive models parametrized as deep neural networks (called, PixelCNN) achieve state-of-the-art results in image density estimation. Although training is fast, sampling is costly, requiring one network evaluation per pixel; O(N) for N pixels. In this talk, he described a parallelized PixelCNN that achieved competitive density estimation and ordered of magnitude speedup - O(log N) sampling instead of O(N) - enabling the practical generation of high-resolution images. Results would be presented on class-conditional image generation, text-to-image synthesis, and action-conditional video generation. In the second part of the talk, he discussed density estimation in the low-data regime as a meta learning problem.
Scott is a research scientist at DeepMind. He completed his PhD with Honglak Lee at the University of Michigan in 2016. His research focuses on deep learning methods for image generation, object detection, imitation learning and program induction.