BriefGPT.xyz
Feb, 2017
卷积神经网络委员会用于黑色素瘤分类，结合传统和专家知识的图像转换数据增强
Increasing Deep Learning Melanoma Classification by Classical And Expert Knowledge Based Image Transforms
HTML
PDF
Cristina Nader Vasconcelos, Bárbara Nader Vasconcelos
TL;DR
本研究以 ISBI 2017 黑色素瘤分类挑战为背景，通过人工增强图像数据和 CNN 模型委员会的方法来提高皮肤癌早期检测的准确率和分类器的稳定性。
Abstract
skin cancer
is a major public health problem, as is the most common type of cancer and represents more than half of cancer diagnoses worldwide.
early detection
influences the outcome of the disease and motivates
→