“Melanoma lesion detection and segmentation using deep region based convolutional neural network and fuzzy C-means clustering”
A deep region-based convolutional neural network (RCNN) precisely detects the multiple affected regions in the form of bounding boxes that simplify localization through fuzzy -mean (FCM) clustering”
The study is focused only on segmentation based on window bounding boxes, not on precise boundaries around the region of interest. Secondly, this did not work on classification
“Multiple skin lesions diagnostics via integrated deep convolutional networks for segmentation and classification”
The authors propose an integrated diagnostic framework that combines a skin lesion boundary segmentation stage and a multiple skin lesion classification stages
The labeled skin lesion images utilized for both the training and testing were still limited in size
“A probabilistic-based deep learning model for skin lesion segmentation”
“The authors employ an efficient mean-field approximate probabilistic inference approach with a fully connected conditional random field that utilizes a Gaussian kernel”
Here they have analyzed and segmented skin lesion images, but no classifying has been done
“Deep learning based skin lesion segmentation and classification of melanoma using support vector machine (SVM)”
“The authors propose a convolutional neural network (CNN) based U-Net algorithm used for segmentation process. After feature extraction, it was fed to various classifiers”
Here the classifying is done using only one unit, and the depth is only up to 4 convolution steps
“Skin lesion segmentation using deep learning with auxiliary task”
“The authors have proposed CNN architecture using auxiliary information along with edge prediction simultaneously with the segmentation task”
The proposed method fails when the foreground region contains a dark area surrounded by an area with light color, whose appearance is more similar to the healthy skin region
