Table 7 Comparison with the state of Art works.
From: An attention based hybrid approach using CNN and BiLSTM for improved skin lesion classification
Reference | Accuracy (%) | Sensitivity (%) | Specificity (%) | Methods and tools | Contribution |
|---|---|---|---|---|---|
89.28 | 81 | 87.16 | Inception-v3, ResNet-50, Inception-ResNet-v2, and DenseNet-201 | Different CNN network integration for segmentation and multiple classification stages | |
87.42 | 97.04 | 97.23 | ResNet152, InceptionResNet-V2, fine-tuning, Euclidean space, L-2 distance, transfer learning, Augmentation, GPU | Classify skin disease of faces using Euclidean space to compute L-2 distance between images | |
91 | 92 | 93 | Inception V3 (GoogleNet), PECK, SCIDOG, SVM, RF, fine-tuning, transfer learning, Augmentation, GPU | Proposing an algorithm that is able to train CNN with limited data | |
87.7 | 85 | 73.29 | AlexNet, VGG, ResNet-18, ResNet-101, SVM, MLP, random forest, transfer learning, Augmentation, GPU | Skin lesion classification using 4 CNNs and ensembling of the final classification results | |
Proposed Model | 92.73 | 92.73 | 98.79 | CNN with BiLSTM and attebntion mechanisms | Improved skin lesion classification |
