39类皮肤病变数据集
收藏资源简介:
该数据集由吉大港工程技术大学和查尔斯特大学的研究团队创建,整合了五个公开可用的数据集,涵盖了39种不同类型的皮肤病变。数据集内容丰富,包含了多种皮肤病变的图像数据,旨在为皮肤病变的分类提供多样化的训练样本。数据集的创建过程包括数据整合、预处理和图像增强等步骤,以提高数据质量和模型的泛化能力。该数据集主要用于深度学习模型的训练和评估,旨在提高皮肤病变分类的准确性和效率,帮助医疗专业人员更快速、准确地诊断多种皮肤病变。
This dataset was developed by a research team from Chittagong University of Engineering & Technology and Charles Sturt University. It integrates five publicly available datasets, covering 39 distinct types of skin lesions. With comprehensive content, this dataset includes image data of multiple skin lesion variants, aiming to provide diverse training samples for skin lesion classification tasks. The development process of this dataset encompasses steps such as data integration, preprocessing, and image augmentation, to improve data quality and the generalization ability of models. This dataset is primarily utilized for training and evaluating deep learning models, with the objective of advancing the accuracy and efficiency of skin lesion classification, thus assisting medical professionals in diagnosing a wide range of skin lesions more rapidly and accurately.

- 1An Attention-Guided Deep Learning Approach for Classifying 39 Skin Lesion Types吉大港工程技术大学计算机科学与工程系, 查尔斯特大学计算、数学与工程学院 · 2025年



