Skin lesion datasets consist predominantly of normal samples with only a small percentage of abnormal ones, giving rise to the class imbalance problem. Also, skin lesion images are largely similar in overall appearance owing to the low inter-class variability. In this paper, we propose a two-stage framework for automatic classification of skin lesion images using adversarial training and transfer learning toward melanoma detection. In the first stage, we leverage the inter-class variation of the data distribution for the task of conditional image synthesis by learning the inter-class mapping and synthesizing under-represented class samples from the over-represented ones using unpaired image-to-image translation. In the second stage, we train a deep convolutional neural network for skin lesion classification using the original training set combined with the newly synthesized under-represented class samples. The training of this classifier is carried out by minimizing the focal loss function, which assists the model in learning from hard examples, while down-weighting the easy ones. Experiments conducted on a dermatology image benchmark demonstrate the superiority of our proposed approach over several standard baseline methods, achieving significant performance improvements. Interestingly, we show through feature visualization and analysis that our method leads to context based lesion assessment that can reach an expert dermatologist level.
Synthetic Melanoma Image Generation and Evaluation Using Generative Adversarial Networks
Department of Engineering Technology, University of Houston（休斯顿大学工程技术系）；Department of Electrical and Computer Engineering, University of Houston（休斯顿大学电气与计算机工程系）；Department of Information Science Technology, University of Houston（休斯顿大学信息科学与技术系）；School of Computing, Clemson University（克莱姆斯大学计算学院）；Department of Public Health, Texas Tech University（德克萨斯技术大学公共卫生系）；Department of Dermatology and Pathology, Texas Tech University（德克萨斯技术大学皮肤科与病理学系）；Department of Dermatology, University of Texas MD Anderson Cancer Center（德克萨斯大学MD安德森癌症中心皮肤科）；Department of Biomedical Engineering, University of Houston（休斯顿大学生物医学工程系）
Unsupervised and semi-supervised learning with Categorical Generative Adversarial Networks assisted by Wasserstein distance for dermoscopy image Classification
University of Saskatchewan（萨斯喀彻温大学）；Philips Canada（加拿大飞利浦）
The Fast and Accurate Approach to Detection and Segmentation of Melanoma Skin Cancer using Fine-tuned Yolov3 and SegNet Based on Deep Transfer Learning
黑色素瘤是最严重的皮肤癌之一，可以发生在人类皮肤的任何部分。黑色素瘤病变的早期诊断将大大增加其治愈的机会。改善黑色素瘤分割将有助于医生或手术机器人更准确地从身体部位切除病变。近年来，基于学习的图像分割方法在图像分割中取得了较好的效果。这项研究提出了一种新的方法，通过定义一个基于深度学习模型的两步管道来改善黑色素瘤皮肤病变的检测和分割。我们的方法在ISIC 2018（皮肤病变分析对黑色素瘤检测挑战数据集）众所周知的数据集上进行了评估。所提出的方法包括两个主要部分的实时检测病变的位置和分割。在检测部分，皮肤病变的位置由微调的You Only Look Once第3版（F-YOLOv 3）精确检测，然后输入微调的分割网络（F-SegNet）。皮肤病变定位有助于减少用于分割的整个图像的不必要的计算。结果表明，我们提出的F-YOLOv 3在平均精度（mAP）方面表现更好，达到96%。与最先进的分割方法相比，我们的F-SegNet实现了更高的性能，准确率为95.16%。
Representation Learning for Non-Melanoma Skin Cancer using a Latent Autoencoder
Australian e-Health Research Centre, CSIRO（澳大利亚联邦科学与工业研究组织电子健康研究中心）
生成学习是表征学习的一个强大工具，在生物医学成像问题上表现出了特别的希望。然而，在这种情况下，从分布中采样是次要的，以找到真实图像的表示，这往往带有标签，并明确表示目标分布的内容和质量。从生成模型忠实地重建图像仍然很困难，特别是那些像组织学图像一样复杂的模型。在这项工作中，结合了两种现有的方法（自动编码器和对抗性潜伏自动编码器），试图提高我们对非黑色素瘤皮肤癌，特别是表皮内癌（IEC）的真实图像进行编码和解码的能力。利用IEC（256 x 256）的高质量图像数据集，这项工作评估了图像重建质量和表示学习的结果。研究表明，对抗性训练可以将基线FID分数从76提高到50，并且表征学习的基准可以提高3%。平滑和现实的插值的形态结构的变化也首次提出，定位表示学习作为一个有前途的方向，在计算病理学的背景下。
GE-AdvGAN: Improving the transferability of adversarial samples by gradient editing-based adversarial generative model
The University of Sydney（悉尼大学）；University of Malaya（马来亚大学）；Suzhou Yierqi（苏州意尔奇）；The University of Texas at San Antonio（德克萨斯大学圣安东尼奥分校）；University of Wollongong（伍伦贡大学）
Protecting Facial Privacy: Generating Adversarial Identity Masks via Style-robust Makeup Transfer
Huazhong University of Science and Technology（华中科技大学）；National Engineering Research Center for Big Data Technology and System（国家大数据技术与系统工程研究中心）；Hubei Engineering Research Center on Big Data Security（湖北省大数据安全工程研究中心）；Deakin University（迪肯大学）
