In this paper, we study the influence of both long and short skip connections on Fully Convolutional Networks (FCN) for biomedical image segmentation. In standard FCNs, only long skip connections are used to skip features from the contracting path to the expanding path in order to recover spatial information lost during downsampling. We extend FCNs by adding short skip connections, that are similar to the ones introduced in residual networks, in order to build very deep FCNs (of hundreds of layers). A review of the gradient flow confirms that for a very deep FCN it is beneficial to have both long and short skip connections. Finally, we show that a very deep FCN can achieve near-to-state-of-the-art results on the EM dataset without any further post-processing.
Description
The Importance of Skip Connections in Biomedical Image Segmentation | SpringerLink
%0 Conference Paper
%1 10.1007/978-3-319-46976-8_19
%A Drozdzal, Michal
%A Vorontsov, Eugene
%A Chartrand, Gabriel
%A Kadoury, Samuel
%A Pal, Chris
%B Deep Learning and Data Labeling for Medical Applications
%C Cham
%D 2016
%E Carneiro, Gustavo
%E Mateus, Diana
%E Peter, Loïc
%E Bradley, Andrew
%E Tavares, João Manuel R. S.
%E Belagiannis, Vasileios
%E Papa, João Paulo
%E Nascimento, Jacinto C.
%E Loog, Marco
%E Lu, Zhi
%E Cardoso, Jaime S.
%E Cornebise, Julien
%I Springer International Publishing
%K ResNet U-net order1
%P 179--187
%T The Importance of Skip Connections in Biomedical Image Segmentation
%X In this paper, we study the influence of both long and short skip connections on Fully Convolutional Networks (FCN) for biomedical image segmentation. In standard FCNs, only long skip connections are used to skip features from the contracting path to the expanding path in order to recover spatial information lost during downsampling. We extend FCNs by adding short skip connections, that are similar to the ones introduced in residual networks, in order to build very deep FCNs (of hundreds of layers). A review of the gradient flow confirms that for a very deep FCN it is beneficial to have both long and short skip connections. Finally, we show that a very deep FCN can achieve near-to-state-of-the-art results on the EM dataset without any further post-processing.
%@ 978-3-319-46976-8
@inproceedings{10.1007/978-3-319-46976-8_19,
abstract = {In this paper, we study the influence of both long and short skip connections on Fully Convolutional Networks (FCN) for biomedical image segmentation. In standard FCNs, only long skip connections are used to skip features from the contracting path to the expanding path in order to recover spatial information lost during downsampling. We extend FCNs by adding short skip connections, that are similar to the ones introduced in residual networks, in order to build very deep FCNs (of hundreds of layers). A review of the gradient flow confirms that for a very deep FCN it is beneficial to have both long and short skip connections. Finally, we show that a very deep FCN can achieve near-to-state-of-the-art results on the EM dataset without any further post-processing.},
added-at = {2020-04-30T20:00:25.000+0200},
address = {Cham},
author = {Drozdzal, Michal and Vorontsov, Eugene and Chartrand, Gabriel and Kadoury, Samuel and Pal, Chris},
biburl = {https://www.bibsonomy.org/bibtex/25f3ac24f47e68a42d0adf3f81b9b7b98/sohnki},
booktitle = {Deep Learning and Data Labeling for Medical Applications},
description = {The Importance of Skip Connections in Biomedical Image Segmentation | SpringerLink},
editor = {Carneiro, Gustavo and Mateus, Diana and Peter, Lo{\"i}c and Bradley, Andrew and Tavares, Jo{\~a}o Manuel R. S. and Belagiannis, Vasileios and Papa, Jo{\~a}o Paulo and Nascimento, Jacinto C. and Loog, Marco and Lu, Zhi and Cardoso, Jaime S. and Cornebise, Julien},
interhash = {a3117a01feed7b2b90b4b1ebddb5572e},
intrahash = {5f3ac24f47e68a42d0adf3f81b9b7b98},
isbn = {978-3-319-46976-8},
keywords = {ResNet U-net order1},
pages = {179--187},
publisher = {Springer International Publishing},
timestamp = {2020-06-02T20:01:26.000+0200},
title = {The Importance of Skip Connections in Biomedical Image Segmentation},
year = 2016
}