Real-time semantic segmentation plays an important role in practical
applications such as self-driving and robots. Most research working on semantic
segmentation focuses on accuracy with little consideration for efficiency.
Several existing studies that emphasize high-speed inference often cannot
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 incorporating the
dilated convolution and the dense connectivity to attain high efficiency at low
computational cost, inference time, and model size. Compared to FCN, EDANet is
11 times faster and has 196 times fewer parameters, while it achieves a higher
the mean of intersection-over-union (mIoU) score without any additional decoder
structure, context module, post-processing scheme, and pretrained model. We
evaluate EDANet on Cityscapes and CamVid datasets to evaluate its performance
and compare it with the other state-of-art systems. Our network can run on
resolution 512x1024 inputs at the speed of 108 and 81 frames per second on a
single GTX 1080Ti and Titan X, respectively.
%0 Generic
%1 citeulike:14653618
%A xxx,
%D 2018
%K arch backbone dilated segmentation
%T Efficient Dense Modules of Asymmetric Convolution for Real-Time Semantic Segmentation
%U http://arxiv.org/abs/1809.06323
%X Real-time semantic segmentation plays an important role in practical
applications such as self-driving and robots. Most research working on semantic
segmentation focuses on accuracy with little consideration for efficiency.
Several existing studies that emphasize high-speed inference often cannot
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 incorporating the
dilated convolution and the dense connectivity to attain high efficiency at low
computational cost, inference time, and model size. Compared to FCN, EDANet is
11 times faster and has 196 times fewer parameters, while it achieves a higher
the mean of intersection-over-union (mIoU) score without any additional decoder
structure, context module, post-processing scheme, and pretrained model. We
evaluate EDANet on Cityscapes and CamVid datasets to evaluate its performance
and compare it with the other state-of-art systems. Our network can run on
resolution 512x1024 inputs at the speed of 108 and 81 frames per second on a
single GTX 1080Ti and Titan X, respectively.
@misc{citeulike:14653618,
abstract = {{ Real-time semantic segmentation plays an important role in practical
applications such as self-driving and robots. Most research working on semantic
segmentation focuses on accuracy with little consideration for efficiency.
Several existing studies that emphasize high-speed inference often cannot
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 incorporating the
dilated convolution and the dense connectivity to attain high efficiency at low
computational cost, inference time, and model size. Compared to FCN, EDANet is
11 times faster and has 196 times fewer parameters, while it achieves a higher
the mean of intersection-over-union (mIoU) score without any additional decoder
structure, context module, post-processing scheme, and pretrained model. We
evaluate EDANet on Cityscapes and CamVid datasets to evaluate its performance
and compare it with the other state-of-art systems. Our network can run on
resolution 512x1024 inputs at the speed of 108 and 81 frames per second on a
single GTX 1080Ti and Titan X, respectively.}},
added-at = {2019-02-27T22:23:29.000+0100},
archiveprefix = {arXiv},
author = {xxx},
biburl = {https://www.bibsonomy.org/bibtex/27a55507b3d8fc03ba2bf608d2b939105/nmatsuk},
citeulike-article-id = {14653618},
citeulike-linkout-0 = {http://arxiv.org/abs/1809.06323},
citeulike-linkout-1 = {http://arxiv.org/pdf/1809.06323},
day = 17,
eprint = {1809.06323},
interhash = {a01dc9146a54dec855f6b35f9c6c8f6a},
intrahash = {7a55507b3d8fc03ba2bf608d2b939105},
keywords = {arch backbone dilated segmentation},
month = sep,
posted-at = {2018-11-08 07:28:49},
priority = {5},
timestamp = {2019-02-27T22:23:29.000+0100},
title = {{Efficient Dense Modules of Asymmetric Convolution for Real-Time Semantic Segmentation}},
url = {http://arxiv.org/abs/1809.06323},
year = 2018
}