Parallel ResNet Fused Deep Learning Model for Skin Cancer Detection

Early detection of skin cancer requires accurate methods to enhance survival statistics because skin cancer persists as a global life-threatening condition. Clinical diagnosis by dermatologists through visual evaluation proves to be subjective and requires extensive time for examinations. This paper develops a deep learning framework which uses dual-branch ResNet50 and ResNet101 architectures to address the performance challenges during feature extraction and classification. TensorFlow’s ImageDataGenerator enables data processing and enhancement before the dataset allocation into separate training and validation portions.