Fig. 2
From: Attention-Enhanced CNNs and transformers for accurate monkeypox and skin disease detection
Graphical Abstract. (a) Data Preprocessing, where the Mpox Skin Lesion Dataset (MSLD v2.0) is used, followed by resizing, normalization, and data augmentation steps to prepare the images for training. (b) Model Architecture Selection and Enhancement, where we choose the EfficientNetB0 architecture and apply coordinate attention to enhance feature extraction, allowing the model to focus on clinically relevant regions. (c) Attention Mechanism, where the attention mechanism is applied to highlight important features, leading to improved model classification accuracy.
