Ensemble Transfer Learning Approach for Monkeypox Detection

The rapid spread of monkeypox (mpox) across multiple nations has made it a major heath challenge. The early detection of monkeypox is crucial for preventing its transmission, particularly in the face of rising epidemics. The proposed method aims to address the problem by creating a model for monkeypox diagnosis using deep learning and image processing approaches. This paper proposes an innovative approach for monkeypox detection using an ensembled transfer learning framework. The proposed model is an ensemble model that includes DenseNet121, InceptionResNetV2, and EfficientNetB3. Each model was separately fine-tuned and trained to make predictions. The final categorization was based on the average of individual model predictions to increase the detection accuracy. The proposed model achieves accuracy, precision, recall, and F1-score of $0.931,0.910,0.930$, and 0.931 respectively. The performance of the proposed model demonstrates that our system detects monkeypox with high accuracy, which makes it a useful tool for prompt and precise diagnosis in clinical situations.