CSU-Net: Contour Semantic Segmentation Self-Enhancement for Human Head Detection
The computer vision community has made tremendous progress in solving a variety of semantic image understanding tasks, such as classification and segmentation. With the advancement of imaging technology and hardware, image semantic segmentation, through the use of deep learning, is among the most co...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2023-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/10004558/ |
_version_ | 1797959577233260544 |
---|---|
author | Mohamed Chouai Petr Dolezel |
author_facet | Mohamed Chouai Petr Dolezel |
author_sort | Mohamed Chouai |
collection | DOAJ |
description | The computer vision community has made tremendous progress in solving a variety of semantic image understanding tasks, such as classification and segmentation. With the advancement of imaging technology and hardware, image semantic segmentation, through the use of deep learning, is among the most common topics which have been worked on in the last decade. However, image semantic segmentation suffers from several drawbacks such as insufficient detection of object boundaries. In this study, we present a new convolutional neural network architecture called CSU-Net that aims to self-enhance the results of semantic segmentation. The proposed model consists of two strongly concatenated encoder-decoder blocks. With this design, we reduced requirements on computing power and memory size to decrease costs and increase the training/prediction speed. This study also demonstrates the advantage of the proposed system for small training data sets. The proposed approach has been implemented on our private dataset, as well as on a publicly available dataset. A comparative analysis was carried out with four popular segmentation models and three other recently introduced architectures to show the efficiency of the proposed system. CSU-Net outperformed the other competing neural networks that we considered for the comparative study. As an example, it succeeded in improving the traditional U-Net result by approximately 50% in mean Intersection over Union (mIoU) for both tested datasets. Based on our experience, the CSU-Net can improve results of semantic segmentation in many applications. |
first_indexed | 2024-04-11T00:34:40Z |
format | Article |
id | doaj.art-3949ce463df741eda8dbbedbf36a74c5 |
institution | Directory Open Access Journal |
issn | 2169-3536 |
language | English |
last_indexed | 2024-04-11T00:34:40Z |
publishDate | 2023-01-01 |
publisher | IEEE |
record_format | Article |
series | IEEE Access |
spelling | doaj.art-3949ce463df741eda8dbbedbf36a74c52023-01-07T00:00:17ZengIEEEIEEE Access2169-35362023-01-011198799910.1109/ACCESS.2022.323341910004558CSU-Net: Contour Semantic Segmentation Self-Enhancement for Human Head DetectionMohamed Chouai0https://orcid.org/0000-0001-5529-8834Petr Dolezel1https://orcid.org/0000-0002-7359-0764Alfred Wegener Institute, Bremerhaven, GermanyFaculty of Electrical Engineering and Informatics, University of Pardubice, Pardubice, Czech RepublicThe computer vision community has made tremendous progress in solving a variety of semantic image understanding tasks, such as classification and segmentation. With the advancement of imaging technology and hardware, image semantic segmentation, through the use of deep learning, is among the most common topics which have been worked on in the last decade. However, image semantic segmentation suffers from several drawbacks such as insufficient detection of object boundaries. In this study, we present a new convolutional neural network architecture called CSU-Net that aims to self-enhance the results of semantic segmentation. The proposed model consists of two strongly concatenated encoder-decoder blocks. With this design, we reduced requirements on computing power and memory size to decrease costs and increase the training/prediction speed. This study also demonstrates the advantage of the proposed system for small training data sets. The proposed approach has been implemented on our private dataset, as well as on a publicly available dataset. A comparative analysis was carried out with four popular segmentation models and three other recently introduced architectures to show the efficiency of the proposed system. CSU-Net outperformed the other competing neural networks that we considered for the comparative study. As an example, it succeeded in improving the traditional U-Net result by approximately 50% in mean Intersection over Union (mIoU) for both tested datasets. Based on our experience, the CSU-Net can improve results of semantic segmentation in many applications.https://ieeexplore.ieee.org/document/10004558/Safety systemshead detectionhead countingsemantic segmentationself-enhancement |
spellingShingle | Mohamed Chouai Petr Dolezel CSU-Net: Contour Semantic Segmentation Self-Enhancement for Human Head Detection IEEE Access Safety systems head detection head counting semantic segmentation self-enhancement |
title | CSU-Net: Contour Semantic Segmentation Self-Enhancement for Human Head Detection |
title_full | CSU-Net: Contour Semantic Segmentation Self-Enhancement for Human Head Detection |
title_fullStr | CSU-Net: Contour Semantic Segmentation Self-Enhancement for Human Head Detection |
title_full_unstemmed | CSU-Net: Contour Semantic Segmentation Self-Enhancement for Human Head Detection |
title_short | CSU-Net: Contour Semantic Segmentation Self-Enhancement for Human Head Detection |
title_sort | csu net contour semantic segmentation self enhancement for human head detection |
topic | Safety systems head detection head counting semantic segmentation self-enhancement |
url | https://ieeexplore.ieee.org/document/10004558/ |
work_keys_str_mv | AT mohamedchouai csunetcontoursemanticsegmentationselfenhancementforhumanheaddetection AT petrdolezel csunetcontoursemanticsegmentationselfenhancementforhumanheaddetection |