Semantic Binary Segmentation using Convolutional Networks without Decoders

2018/05/01 by Shubhra Aich, Aich, Shubhra, William van der Kamp +3
Computer Science · #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #cs.CV

paper · pdf · doi:10.48550/arxiv.1805.00138

CVPR 2018 DeepGlobe Workshop; Code repository: https://github.com/littleaich/deepglobe2018

arxiv created 2018/05/25 · arxiv updated 2018/05/29

Abstract

In this paper, we propose an efficient architecture for semantic image segmentation using the depth-to-space (D2S) operation. Our D2S model is comprised of a standard CNN encoder followed by a depth-to-space reordering of the final convolutional feature maps. Our approach eliminates the decoder portion of traditional encoder-decoder segmentation models and reduces the amount of computation almost by half. As a participant of the DeepGlobe Road Extraction competition, we evaluate our models on the corresponding road segmentation dataset. Our highly efficient D2S models exhibit comparable performance to standard segmentation models with much lower computational cost.

Related

• CNN-aware Binary Map for General Semantic Segmentation
• Efficient Convolutional Neural Network with Binary Quantization Layer
• High-performance Semantic Segmentation Using Very Deep Fully Convolutional Networks
• Beyond Single Stage Encoder-Decoder Networks: Deep Decoders for Semantic Image Segmentation
• Real time backbone for semantic segmentation