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) -
LA-ResUNet: An Efficient Linear Attention Mechanism in ResUNet for the Semantic Segmentation of Pulmonary Nodules
by: P. C. Sarah Prithvika, et al.
Published: (2024-01-01) -
CLM-UNet: A Remote Sensing Image Semantic Segmentation Method Incorporating Attention Mechanism
by: Panpan Zhang, et al.
Published: (2025-01-01) -
GCBAM-UNet: Sun Glare Segmentation Using Convolutional Block Attention Module
by: Nabila Zrira, et al.
Published: (2024-12-01) -
DAF-UNet: Deformable U-Net with Atrous-Convolution Feature Pyramid for Retinal Vessel Segmentation
by: Yongchao Duan, et al.
Published: (2025-04-01)