This repo is an implementation of network described in Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation. DeepLab is a deep learning model for semantic image segmentation.
The implementation of the authors of the paper can be found here here.