About this document
Efficient Deep Learning For Skin Lesion Segmentation Using Convolutional Block Attention Mechanism by Rayed Ahmed is a document available to read on EtoBox.
The thesis presents a hybrid deep learning model for automated skin lesion segmentation to enhance early diagnosis of skin cancer, particularly melanoma. Utilizing the HAM10000 dataset, the model integrates EfficientNetB7 and UNet++ with Convolutional Block Attention Modules, achieving notable performance improvements in segmentation metrics. Future work aims to refine preprocessing techniques and incorporate multimodal data to further advance dermatological care.
- Author
- Rayed Ahmed
- Language
- EN