icnet

ICNet

ICNet for Real-time Semantic Segmentation on High-resolution Images

Introduction

Official Repo

Code Snippet

Abstract

We focus on the challenging task of real-time semantic segmentation in this paper. It finds many practical applications and yet is with fundamental difficulty of reducing a large portion of computation for pixel-wise label inference. We propose an image cascade network (ICNet) that incorporates multi-resolution branches under proper label guidance to address this challenge. We provide in-depth analysis of our framework and introduce the cascade feature fusion unit to quickly achieve high-quality segmentation. Our system yields real-time inference on a single GPU card with decent quality results evaluated on challenging datasets like Cityscapes, CamVid and COCO-Stuff.

Citation

@inproceedings{zhao2018icnet,
  title={Icnet for real-time semantic segmentation on high-resolution images},
  author={Zhao, Hengshuang and Qi, Xiaojuan and Shen, Xiaoyong and Shi, Jianping and Jia, Jiaya},
  booktitle={Proceedings of the European conference on computer vision (ECCV)},
  pages={405--420},
  year={2018}
}

Results and models

Cityscapes

Method	Backbone	Crop Size	Lr schd	Mem (GB)	Inf time (fps)	mIoU	mIoU(ms+flip)	config	download
ICNet	R-18-D8	832x832	80000	1.70	27.12	68.14	70.16	config	model | log
ICNet	R-18-D8	832x832	160000	-	-	71.64	74.18	config	model | log
ICNet (in1k-pre)	R-18-D8	832x832	80000	-	-	72.51	74.78	config	model | log
ICNet (in1k-pre)	R-18-D8	832x832	160000	-	-	74.43	76.72	config	model | log
ICNet	R-50-D8	832x832	80000	2.53	20.08	68.91	69.72	config	model | log
ICNet	R-50-D8	832x832	160000	-	-	73.82	75.67	config	model | log
ICNet (in1k-pre)	R-50-D8	832x832	80000	-	-	74.58	76.41	config	model | log
ICNet (in1k-pre)	R-50-D8	832x832	160000	-	-	76.29	78.09	config	model | log
ICNet	R-101-D8	832x832	80000	3.08	16.95	70.28	71.95	config	model | log
ICNet	R-101-D8	832x832	160000	-	-	73.80	76.10	config	model | log
ICNet (in1k-pre)	R-101-D8	832x832	80000	-	-	75.57	77.86	config	model | log
ICNet (in1k-pre)	R-101-D8	832x832	160000	-	-	76.15	77.98	config	model | log

Note: in1k-pre means pretrained model is used.