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...
Saved in:
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!
|
Similar Items
-
Advancing semantic segmentation: Enhanced UNet algorithm with attention mechanism and deformable convolution
by: Effat Sahragard, et al.
Published: (2025-01-01) -
DS-SwinUNet: Redesigning Skip Connection With Double Scale Attention for Land Cover Semantic Segmentation
by: Zirui Shen, et al.
Published: (2025-01-01) -
D2CBDAMAttUnet: Dual-Decoder Convolution Block Dual Attention Unet for Accurate Retinal Vessel Segmentation From Fundus Images
by: Vo Trong Quang Huy, et al.
Published: (2025-01-01) -
Skin Lesion Image Segmentation Algorithm Based on MC-UNet
by: Guihua Yang, et al.
Published: (2025-01-01) -
Improved lung nodule segmentation with a squeeze excitation dilated attention based residual UNet
by: Dhafer Alhajim, et al.
Published: (2025-01-01)