Loading KodeToCareer...
A CNN computer vision deep learning model detecting pneumonia indicators from chest X-Ray image uploads with Grad-CAM.
This project utilizes deep convolutional neural networks (CNNs) trained on chest X-Ray datasets. It classifies images into normal or bacterial/viral pneumonia indicators, showing heatmaps of detected anomaly areas to assist medical analysts.
Handling dataset imbalance where viral pneumonia samples were significantly fewer than normal samples.
Implemented data augmentation techniques including zoom, shear, and brightness shifting inside the training pipeline.
Achieved 96.2% diagnostic accuracy on the validation dataset.
Utilized Grad-CAM to highlight classification visual decision regions in x-ray uploads.
Created an interactive web app with Streamlit for doctor-facing analysis.
Learn the exact tools, architectures, and programming systems used by Priya Iyer to deploy this project. Join our next live career cohort.
