Deep Optical Flow Learning Networks Combined with Attention Mechanism
In order to improve the accuracy of deep learning optical flow estimation based on encoder-decoder U-Net, a modified supervised deep optical flow learning network combined with attention mechanism is proposed, which consists of a contracting part and an expanding part. In contracting part, high-leve...
Main Author: | |
---|---|
Format: | Article |
Language: | zho |
Published: |
Journal of Computer Engineering and Applications Beijing Co., Ltd., Science Press
2020-11-01
|
Series: | Jisuanji kexue yu tansuo |
Subjects: | |
Online Access: | http://fcst.ceaj.org/CN/abstract/abstract2444.shtml |
id |
doaj-576b5e78f89c487e9422567f79492dfd |
---|---|
record_format |
Article |
spelling |
doaj-576b5e78f89c487e9422567f79492dfd2021-08-10T09:07:04ZzhoJournal of Computer Engineering and Applications Beijing Co., Ltd., Science PressJisuanji kexue yu tansuo1673-94182020-11-0114111920192910.3778/j.issn.1673-9418.1910052Deep Optical Flow Learning Networks Combined with Attention MechanismZHOU Haiyun, XIANG Xuezhi, ZHAI Mingliang, ZHANG Rongfang, WANG Shuai01. Institute of Public Security, Nanjing Forest Police College, Nanjing 210023, China 2. School of Information and Communication Engineering, Harbin Engineering University, Harbin 150001, ChinaIn order to improve the accuracy of deep learning optical flow estimation based on encoder-decoder U-Net, a modified supervised deep optical flow learning network combined with attention mechanism is proposed, which consists of a contracting part and an expanding part. In contracting part, high-level feature information is ex-tracted using a series of convolutional layers, and spatial feature maps are then restored to full resolution by conducting successive deconvolution in expanding part. In this paper, attention mechanism is embedded in U-Net to learn inter-dependencies among the channels so that the channel-wise features can be weighted adaptively, which can enhance the performance of feature extraction. Meanwhile, the proposed network also combines dilated convolution to enlarge the receptive field without changing the size of convolutional kernel. Further, constancy constraints and smoothness constraints from variational method are also adopted so that priori knowledge can be used to improve the accuracy of optical flow estimation. Extensive experiments are conducted on synthesis image sequence datasets and the experi-mental results show the proposed network is effective for improving accuracy of deep learning optical flow estimation.http://fcst.ceaj.org/CN/abstract/abstract2444.shtmloptical flow estimationdeep learningattention mechanismdilated convolutionprior constraints |
collection |
DOAJ |
language |
zho |
format |
Article |
sources |
DOAJ |
author |
ZHOU Haiyun, XIANG Xuezhi, ZHAI Mingliang, ZHANG Rongfang, WANG Shuai |
spellingShingle |
ZHOU Haiyun, XIANG Xuezhi, ZHAI Mingliang, ZHANG Rongfang, WANG Shuai Deep Optical Flow Learning Networks Combined with Attention Mechanism Jisuanji kexue yu tansuo optical flow estimation deep learning attention mechanism dilated convolution prior constraints |
author_facet |
ZHOU Haiyun, XIANG Xuezhi, ZHAI Mingliang, ZHANG Rongfang, WANG Shuai |
author_sort |
ZHOU Haiyun, XIANG Xuezhi, ZHAI Mingliang, ZHANG Rongfang, WANG Shuai |
title |
Deep Optical Flow Learning Networks Combined with Attention Mechanism |
title_short |
Deep Optical Flow Learning Networks Combined with Attention Mechanism |
title_full |
Deep Optical Flow Learning Networks Combined with Attention Mechanism |
title_fullStr |
Deep Optical Flow Learning Networks Combined with Attention Mechanism |
title_full_unstemmed |
Deep Optical Flow Learning Networks Combined with Attention Mechanism |
title_sort |
deep optical flow learning networks combined with attention mechanism |
publisher |
Journal of Computer Engineering and Applications Beijing Co., Ltd., Science Press |
series |
Jisuanji kexue yu tansuo |
issn |
1673-9418 |
publishDate |
2020-11-01 |
description |
In order to improve the accuracy of deep learning optical flow estimation based on encoder-decoder U-Net, a modified supervised deep optical flow learning network combined with attention mechanism is proposed, which consists of a contracting part and an expanding part. In contracting part, high-level feature information is ex-tracted using a series of convolutional layers, and spatial feature maps are then restored to full resolution by conducting successive deconvolution in expanding part. In this paper, attention mechanism is embedded in U-Net to learn inter-dependencies among the channels so that the channel-wise features can be weighted adaptively, which can enhance the performance of feature extraction. Meanwhile, the proposed network also combines dilated convolution to enlarge the receptive field without changing the size of convolutional kernel. Further, constancy constraints and smoothness constraints from variational method are also adopted so that priori knowledge can be used to improve the accuracy of optical flow estimation. Extensive experiments are conducted on synthesis image sequence datasets and the experi-mental results show the proposed network is effective for improving accuracy of deep learning optical flow estimation. |
topic |
optical flow estimation deep learning attention mechanism dilated convolution prior constraints |
url |
http://fcst.ceaj.org/CN/abstract/abstract2444.shtml |
work_keys_str_mv |
AT zhouhaiyunxiangxuezhizhaimingliangzhangrongfangwangshuai deepopticalflowlearningnetworkscombinedwithattentionmechanism |
_version_ |
1721212361128280064 |