Draw: A recurrent neural network for image generation
K Gregor, I Danihelka, A Graves… - International …, 2015 - proceedings.mlr.press
International conference on machine learning, 2015•proceedings.mlr.press
This paper introduces the Deep Recurrent Attentive Writer (DRAW) architecture for image
generation with neural networks. DRAW networks combine a novel spatial attention
mechanism that mimics the foveation of the human eye, with a sequential variational auto-
encoding framework that allows for the iterative construction of complex images. The system
substantially improves on the state of the art for generative models on MNIST, and, when
trained on the Street View House Numbers dataset, it is able to generate images that are …
generation with neural networks. DRAW networks combine a novel spatial attention
mechanism that mimics the foveation of the human eye, with a sequential variational auto-
encoding framework that allows for the iterative construction of complex images. The system
substantially improves on the state of the art for generative models on MNIST, and, when
trained on the Street View House Numbers dataset, it is able to generate images that are …
Abstract
This paper introduces the Deep Recurrent Attentive Writer (DRAW) architecture for image generation with neural networks. DRAW networks combine a novel spatial attention mechanism that mimics the foveation of the human eye, with a sequential variational auto-encoding framework that allows for the iterative construction of complex images. The system substantially improves on the state of the art for generative models on MNIST, and, when trained on the Street View House Numbers dataset, it is able to generate images that are indistinguishable from real data with the naked eye.
proceedings.mlr.press