Unsupervised and semi-supervised learning with categorical generative adversarial networks assisted by wasserstein distance for dermoscopy image classification

X Yi, E Walia, P Babyn - arXiv preprint arXiv:1804.03700, 2018 - arxiv.org
… feature representation of dermoscopy images in an unsupervised and semi-supervised
manner. … images. Moreover, the proposed method is also capable of generating real-world like …

Dynamic clustering detection through multi‐valued descriptors of dermoscopic images

V Cozza, MR Guarracino, L Maddalena… - Statistics in …, 2011 - Wiley Online Library
… , a segmentation of dermoscopic images allows to identify a … a set of images where there is
an a priori classification so that it … propose a distance based on the Wasserstein metric 15 for …

Skin lesion classification by multi-view filtered transfer learning

J Bian, S Zhang, S Wang, J Zhang, J Guo - IEEE Access, 2021 - ieeexplore.ieee.org
… Pm by minimizing the Wasserstein distance between target image features as the source ones.
… , “A methodological approach to the classification of dermoscopy images,” Computerized …

Few-Shot Classification with Multiscale Feature Fusion for Clinical Skin Disease Diagnosis

T Chen, Q Liu, J Yang - Clinical, Cosmetic and Investigational …, 2024 - Taylor & Francis
… of publicly accessible dermoscopic images, Citation36 which … benchmark competition for
dermoscopic image analysis. … image consists of a dermoscopic image and a clinical image

GAN based augmentation using a hybrid loss function for dermoscopy images

E Goceri - Artificial Intelligence Review, 2024 - Springer
… Therefore, in this work, the effectiveness of those GAN-based augmentation techniques
for dermoscopy images and their significance in classification has been presented by fair …

Self-supervised wasserstein pseudo-labeling for semi-supervised image classification

F Taherkhani, A Dabouei… - … pattern recognition, 2021 - openaccess.thecvf.com
… In this work, we propose a new SSL method based on the Wasserstein metric which follows
the general assumptions in SSL. Inspired by the effectiveness of Self-Supervised learning in …

Generating highly realistic images of skin lesions with GANs

C Baur, S Albarqouni, N Navab - … Robotic Endoscopy, Clinical Image …, 2018 - Springer
… For our experiments, we utilize dermoscopic images of benign … sample realism, the Sliced
Wasserstein Distance (SWD) has … these 80 images in random order and let him/her classify. …

Classification of skin lesions by Tda alongside Xception neural network

N Elyasi, M Hosseini Moghadam - Journal of AI and Data …, 2022 - jad.shahroodut.ac.ir
… In [20] authors used a dataset of 13000 dermoscopic images, including HAM10000 which
… We will bring the results of computing Wasserstein distances for different epochs in Table 5. …

MelanoGANs: high resolution skin lesion synthesis with GANs

C Baur, S Albarqouni, N Navab - arXiv preprint arXiv:1804.04338, 2018 - arxiv.org
… experiments for skin lesion classification we further show that … consisting of 2000 dermoscopic
images of both benign and … Wasserstein Distance between the histogram of the training …

A GAN-based image synthesis method for skin lesion classification

Z Qin, Z Liu, P Zhu, Y Xue - Computer Methods and Programs in …, 2020 - Elsevier
… This research benefits from the high-quality dermoscopy images provided by the International
Skin Imaging Collaboration (ISIC). ISIC has made great contributions to the computer-…