RepVGG-SE: An Enhanced Deep Learning Architecture with Squeeze-and-Excitation Attention for Automated Skin Cancer Detection

DOI

10.2991/978-94-6463-940-7_33How to use a DOI?

Keywords

Basal Cell Carcinoma; Deep Learning; Skin Cancer Detection; RepVGG; Squeeze-and-Excitation Attention; Medical Image Classification

Abstract

BCC is among the most prevalent skin cancers, making timely diagnosis essential for effective treatment and tissue preservation. This study introduces a deep learning method for automatically classifying dermoscopic images as either BCC or healthy, utilizing a modified RepVGG-A0 network combined with Squeeze-and-Excitation (SE) blocks. The model is lightweight, quick, and designed for easy use in clinical settings. The dataset includes labeled dermoscopic images divided into “bcc” and “healthy” classes. We applied several preprocessing and augmentation techniques, like random cropping, rotation, flipping, and color adjustments, to improve the model’s performance. The RepVGG backbone offers strong hierarchical feature extraction, while the SE blocks recalibrate feature maps channel-by-channel to boost learning. The network was trained for over 60 epochs using the Adam optimizer and a cross-entropy loss function. We assessed performance using metrics such as accuracy, precision, recall, F1-score, confusion matrix, and ROC-AUC. The model achieved a test accuracy of 98.43%, precision of 98.66%, recall of 98.44%, and an F1-score of 98.55%, showing strong classification performance. We also created a prediction interface to classify unseen images in real-time. This interface shows both the predicted label and confidence score along with the image. This system has great potential for skin cancer screening tools. The SE-enhanced RepVGG allows for accurate and quick classification, making it suitable for integration into clinical workflows or mobile diagnostic applications.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Sai Tummala
AU  - G. Kumari
PY  - 2025
DA  - 2025/12/31
TI  - RepVGG-SE: An Enhanced Deep Learning Architecture with Squeeze-and-Excitation Attention for Automated Skin Cancer Detection
BT  - Proceedings of the Conference on Social and Sustainable Innovation in Technology & Engineering (SASI-ITE 2025)
PB  - Atlantis Press
SP  - 451
EP  - 471
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6463-940-7_33
DO  - 10.2991/978-94-6463-940-7_33
ID  - Tummala2025
ER  -