Oral Cancer Screening App

Deployment of a two-stage model for oral lesion detection and classification with Streamlit, using:

• YOLOv5 for lesion detection
• EfficientNet-b4 for lesion classification

Lesions are classified into 3 categories as benign, precancerous, or cancerous.

Demo

Built With

• [Streamlit] (https://www.streamlit.io/)
• [YOLOv5] (https://github.com/ultralytics/yolov5)
• [EfficientNet] (https://github.com/lukemelas/EfficientNet-PyTorch)