Advancing semantic segmentation: Enhanced UNet algorithm with attention mechanism and deformable convolution.

This paper presents a novel method for improving semantic segmentation performance in computer vision tasks. Our approach utilizes an enhanced UNet architecture that leverages an improved ResNet50 backbone. We replace the last layer of ResNet50 with deformable convolution to enhance feature represen...

Full description

Saved in:
Bibliographic Details
Main Authors: Effat Sahragard, Hassan Farsi, Sajad Mohamadzadeh
Format: Article
Language:English
Published: Public Library of Science (PLoS) 2025-01-01
Series:PLoS ONE
Online Access:https://doi.org/10.1371/journal.pone.0305561
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1832540295761231872
author Effat Sahragard
Hassan Farsi
Sajad Mohamadzadeh
author_facet Effat Sahragard
Hassan Farsi
Sajad Mohamadzadeh
author_sort Effat Sahragard
collection DOAJ
description This paper presents a novel method for improving semantic segmentation performance in computer vision tasks. Our approach utilizes an enhanced UNet architecture that leverages an improved ResNet50 backbone. We replace the last layer of ResNet50 with deformable convolution to enhance feature representation. Additionally, we incorporate an attention mechanism, specifically ECA-ASPP (Attention Spatial Pyramid Pooling), in the encoding path of UNet to capture multi-scale contextual information effectively. In the decoding path of UNet, we explore the use of attention mechanisms after concatenating low-level features with high-level features. Specifically, we investigate two types of attention mechanisms: ECA (Efficient Channel Attention) and LKA (Large Kernel Attention). Our experiments demonstrate that incorporating attention after concatenation improves segmentation accuracy. Furthermore, we compare the performance of ECA and LKA modules in the decoder path. The results indicate that the LKA module outperforms the ECA module. This finding highlights the importance of exploring different attention mechanisms and their impact on segmentation performance. To evaluate the effectiveness of the proposed method, we conduct experiments on benchmark datasets, including Stanford and Cityscapes, as well as the newly introduced WildPASS and DensPASS datasets. Based on our experiments, the proposed method achieved state-of-the-art results including mIoU 85.79 and 82.25 for the Stanford dataset, and the Cityscapes dataset, respectively. The results demonstrate that our proposed method performs well on these datasets, achieving state-of-the-art results with high segmentation accuracy.
format Article
id doaj-art-fb600cc669404c1aa967ced85ae7bee7
institution Kabale University
issn 1932-6203
language English
publishDate 2025-01-01
publisher Public Library of Science (PLoS)
record_format Article
series PLoS ONE
spelling doaj-art-fb600cc669404c1aa967ced85ae7bee72025-02-05T05:31:18ZengPublic Library of Science (PLoS)PLoS ONE1932-62032025-01-01201e030556110.1371/journal.pone.0305561Advancing semantic segmentation: Enhanced UNet algorithm with attention mechanism and deformable convolution.Effat SahragardHassan FarsiSajad MohamadzadehThis paper presents a novel method for improving semantic segmentation performance in computer vision tasks. Our approach utilizes an enhanced UNet architecture that leverages an improved ResNet50 backbone. We replace the last layer of ResNet50 with deformable convolution to enhance feature representation. Additionally, we incorporate an attention mechanism, specifically ECA-ASPP (Attention Spatial Pyramid Pooling), in the encoding path of UNet to capture multi-scale contextual information effectively. In the decoding path of UNet, we explore the use of attention mechanisms after concatenating low-level features with high-level features. Specifically, we investigate two types of attention mechanisms: ECA (Efficient Channel Attention) and LKA (Large Kernel Attention). Our experiments demonstrate that incorporating attention after concatenation improves segmentation accuracy. Furthermore, we compare the performance of ECA and LKA modules in the decoder path. The results indicate that the LKA module outperforms the ECA module. This finding highlights the importance of exploring different attention mechanisms and their impact on segmentation performance. To evaluate the effectiveness of the proposed method, we conduct experiments on benchmark datasets, including Stanford and Cityscapes, as well as the newly introduced WildPASS and DensPASS datasets. Based on our experiments, the proposed method achieved state-of-the-art results including mIoU 85.79 and 82.25 for the Stanford dataset, and the Cityscapes dataset, respectively. The results demonstrate that our proposed method performs well on these datasets, achieving state-of-the-art results with high segmentation accuracy.https://doi.org/10.1371/journal.pone.0305561
spellingShingle Effat Sahragard
Hassan Farsi
Sajad Mohamadzadeh
Advancing semantic segmentation: Enhanced UNet algorithm with attention mechanism and deformable convolution.
PLoS ONE
title Advancing semantic segmentation: Enhanced UNet algorithm with attention mechanism and deformable convolution.
title_full Advancing semantic segmentation: Enhanced UNet algorithm with attention mechanism and deformable convolution.
title_fullStr Advancing semantic segmentation: Enhanced UNet algorithm with attention mechanism and deformable convolution.
title_full_unstemmed Advancing semantic segmentation: Enhanced UNet algorithm with attention mechanism and deformable convolution.
title_short Advancing semantic segmentation: Enhanced UNet algorithm with attention mechanism and deformable convolution.
title_sort advancing semantic segmentation enhanced unet algorithm with attention mechanism and deformable convolution
url https://doi.org/10.1371/journal.pone.0305561
work_keys_str_mv AT effatsahragard advancingsemanticsegmentationenhancedunetalgorithmwithattentionmechanismanddeformableconvolution
AT hassanfarsi advancingsemanticsegmentationenhancedunetalgorithmwithattentionmechanismanddeformableconvolution
AT sajadmohamadzadeh advancingsemanticsegmentationenhancedunetalgorithmwithattentionmechanismanddeformableconvolution