Method Details

Details for method 'ContextNet'

name	ContextNet
challenge	pixel-level semantic labeling
details	Modern deep learning architectures produce highly accurate results on many challenging semantic segmentation datasets. State-of-the-art methods are, however, not directly transferable to real-time applications or embedded devices, since naive adaptation of such systems to reduce computational cost (speed, memory and energy) causes a significant drop in accuracy. We propose ContextNet, a new deep neural network architecture which builds on factorized convolution, network compression and pyramid representations to produce competitive semantic segmentation in real-time with low memory requirements. ContextNet combines a deep branch at low resolution that captures global context information efficiently with a shallow branch that focuses on high-resolution segmentation details. We analyze our network in a thorough ablation study and present results on the Cityscapes dataset, achieving 66.1% accuracy at 18.3 frames per second at full (1024x2048) resolution.
publication	ContextNet: Exploring Context and Detail for Semantic Segmentation in Real-time Rudra PK Poudel, Ujwal Bonde, Stephan Liwicki, Christopher Zach arXiv https://arxiv.org/abs/1805.04554
project page / code
used Cityscapes data	fine annotations
used external data
runtime	0.0238 s Nvidia Titan X (Maxwell, 3,072 CUDA cores)
subsampling	no
submission date	May, 2018
previous submissions