Diabetic foot ulcers (D FU) pose a significant health risk to people with diabetes, requiring early and accurate detection for rapid intervention. In the context of this research, we present a system for automated detection employing (CNN) deep convolutional neural network (CNN) in Efficient Network architecture. This model leverages the power of transfer learning by initializing with pre-trained weights from ImageNet, allowing the model to learn relevant features from a wide variety of images. A dataset consisting of positive samples (DFU) and negative samples (healthy feet) is collected and preprocessed for training. Through customization and fine-tuning of the Performance Network model, we enabled the model to distinguish important features associated with foot ulcers during exercise. Although further validation and collaboration with healthcare experts is necessary, this deep CNN Efficiency Network model shows significant potential as a supporting tool to help improve detection and manage diabetic foot ulcers, thereby reducing the risk of complications associated with the disease.