Real-time perception pipeline for edge devices. Integrates classical OpenCV lane detection with YOLOv8-nano for multi-class object detection on CPU-only hardware.
AI-driven medical assistant utilizing Tesseract-based OCR for document digitization and modular ML models for automated symptom analysis and diagnosis support.
Hybrid CNN-LSTM deep learning architecture for classifying audio into mood categories. Extracts MFCC and spectral features to map audio to the Russell Emotion Model.
Deep learning screening system using MobileNetV2 Transfer Learning to classify Pap smear cell images into distinct diagnostic categories with high precision.
