Convolutional Neural Networks for Medical Image Analysis: Full Training or Fine Tuning?
IEEE Transactions on Medical Imaging(2017)
摘要
Training a deep convolutional neural network (CNN) from scratch is difficult because it requires a large amount of labeled training data and a great deal of expertise to ensure proper convergence. A promising alternative is to fine-tune a CNN that has been pre-trained using, for instance, a large set of labeled natural images. However, the substantial differences between natural and medical images...
更多查看译文
关键词
Biomedical imaging,Training,Feature extraction,Image segmentation,Computed tomography,Image analysis,Tuning
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络