Skip to content

Latest commit

 

History

History
14 lines (10 loc) · 1.01 KB

File metadata and controls

14 lines (10 loc) · 1.01 KB

DRAW: A Recurrent Neural Network For Image Generation

Authors: Karol Gregor, Ivo Danihelka, Alex Graves, Danilo Jimenez Rezende, Daan Wierstra

Problem Statement & Main Contributions

  • The paper presents a sequential VAE with each cell as an LSTM, with the decoder incrementally adding the outputs with a attention update mechanism for restricting the input area by the encoder and the output update area by the decoder.

  • The model is tested on MNIST ,SVHN and CIFAR0-10 dataset.

Related Work (Key Papers)

  • Previously proposed ideas are based either on VAEs or on generative modelling approach which generate images in a single pass.

Assumptions and Future Work

  • The model could have been tested against the performance of GANs and other generative models available. Also, the model when tested against the CIFAR-10 dataset doesnot seem to perform upto mark. These could be some of the potential areas for future work.