Pneumonia Detection (CNN)
Python
PyTorch
Computer Vision
Project Overview
This project involved building a deep learning model to automate the detection of pneumonia from chest X-ray images. By utilizing a pre-trained ResNet-18 architecture and fine-tuning it on the RSNA dataset, I achieved a validation accuracy of 92%.
Key Features
- Implemented custom data augmentation to handle class imbalance.
- Used Grad-CAM to visualize which parts of the X-ray the model focused on.
- Deployed as a microservice using FastAPI.