Random Forest, k-means, SVM, and ResNet

Geometrical templates of gradient vector flow (GVF) and active contours are techniques used for boundary segmentation of skin cancer.

Segmentation and Classification

modified Segmentation-based Fractal (SFTA), lightweight version of the S-MobileNet model, End-to-end deep convolutional neural network

RCNN and FCM clustering

Automatic semantics uses FCN-AlexNet, FCN-8s, FCN-16s, and FCN-32s.

DCNN, Efficient Channel and Spatial Attention Module, Multi-Scale Pyramid Module (MSPM), Pyramid Attention Module

Self-attention based PGAN

Deep learning full resolution convolutional network (FRCN)

Multi-Scale Attention U-Net (MSAU-Net)

Proposed a novel CNN (high-resolution convolutional neural network)

Improved fully convolutional network (IFCN)

Combines two architectures, U-Net with ResNet, collectively called Res-Unet.

Fully automated deep learning ensemble methods depend on Mask R-CNN and DeeplabV3+methods.

Dual-path Feature Fusion (DFF) Module, Attention mechanism

LinkNet-B7

Coupled CNNs creating a triangle map

Parallel Vision Mamba Layer (PVM Layer)

Efficient pyramid squeeze attention (EPSA),Atrous spatial pyramid pooling (ASPP) Multi-layer fusion (MLF)

U-net, Attention modules, Hybrid Features Extraction

deep learning based ‘You Only Look Once (YOLO)’ algorithm

Parallel Network Architecture, U-Net, Atrous Spatial Pyramid Pooling