Related Experiment Video
Updated: Jan 5, 2026
03:31
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
982
Dense pooling layers in fully convolutional network for skin lesion segmentation
Ebrahim Nasr-Esfahani1, Shima Rafiei1, Mohammad H Jafari2
1Department of Electrical and Computer Engineering, Isfahan University of Technology, Isfahan, 84156-83111, Iran.
Summary
A novel deep learning network with dense pooling layers improves skin lesion segmentation accuracy. This method enhances border detection, outperforming current algorithms for computerized skin cancer detection.
Area of Science:
- Medical Image Analysis
- Computer Vision
- Dermatology
Background:
- Accurate lesion segmentation and border detection are crucial for computerized skin cancer detection.
- Existing segmentation methods often struggle with precise border delineation in skin images.
Purpose of the Study:
- To propose a new fully convolutional network architecture for improved skin lesion segmentation.
- To enhance the accuracy of border detection in medical image analysis, specifically for skin lesions.
Main Methods:
- Development of a novel fully convolutional network incorporating new dense pooling layers.
- Application of the proposed network to skin lesion datasets for segmentation tasks.
Main Results:
- The proposed network achieves highly accurate segmentation of skin lesions.
- The method demonstrates superior performance compared to state-of-the-art algorithms in skin lesion segmentation tasks.
Conclusions:
- The novel network with dense pooling layers offers a significant advancement in skin lesion segmentation.
- This approach holds promise for improving the accuracy of automated skin cancer detection systems.
