Things I've built
Developed a Deep Learning model for early-stage breast cancer detection using the ALL-MIAS mammography dataset. Engineered a CNN-based image classification pipeline for automated cancer diagnosis. Achieved 96% test accuracy in classifying mammogram images.
Built a real-time action detection system to classify actions as signed or unsigned from live input streams. Implemented a CNN-based architecture and deployed an interactive Streamlit application for real-time inference. Attained 94% accuracy on evaluation data.