Efficient dense modules of asymmetric convolution for real-time semantic segmentation

Shao Yuan Lo, Hsueh Ming Hang, Sheng Wei Chan, Jing Jhih Li

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Real-time semantic segmentation plays an important role in practical applications such as self-driving and robots. Most semantic segmentation research focuses on improving estimation accuracy with little consideration on efficiency. Several previous studies that emphasize high-speed inference often fail to produce high-accuracy segmentation results. In this paper, we propose a novel convolutional network named Efficient Dense modules with Asymmetric convolution (EDANet), which employs an asymmetric convolution structure and incorporates dilated convolution and dense connectivity to achieve high efficiency at low computational cost and model size. EDANet is 2.7 times faster than the existing fast segmentation network, ICNet, while it achieves a similar mIoU score without any additional context module, post-processing scheme, and pretrained model. We evaluate EDANet on Cityscapes and CamVid datasets, and compare it with the other state-of-art systems. Our network can run with the high-resolution inputs at the speed of 108 FPS on one GTX 1080Ti.

Original languageEnglish
Title of host publication1st ACM International Conference on Multimedia in Asia, MMAsia 2019
PublisherAssociation for Computing Machinery, Inc
ISBN (Electronic)9781450368414
DOIs
StatePublished - 15 Dec 2019
Event1st ACM International Conference on Multimedia in Asia, MMAsia 2019 - Beijing, China
Duration: 15 Dec 201918 Dec 2019

Publication series

Name1st ACM International Conference on Multimedia in Asia, MMAsia 2019

Conference

Conference1st ACM International Conference on Multimedia in Asia, MMAsia 2019
CountryChina
CityBeijing
Period15/12/1918/12/19

Keywords

  • Fast network design
  • Real-time
  • Semantic segmentation

Fingerprint Dive into the research topics of 'Efficient dense modules of asymmetric convolution for real-time semantic segmentation'. Together they form a unique fingerprint.

  • Cite this

    Lo, S. Y., Hang, H. M., Chan, S. W., & Li, J. J. (2019). Efficient dense modules of asymmetric convolution for real-time semantic segmentation. In 1st ACM International Conference on Multimedia in Asia, MMAsia 2019 (1st ACM International Conference on Multimedia in Asia, MMAsia 2019). Association for Computing Machinery, Inc. https://doi.org/10.1145/3338533.3366558