Fake Currency Detection (CNN)
- Goal: Real-time counterfeit detection via web upload.
- Stack: HTML/CSS, Flask, TensorFlow Lite.
- Pipeline: Grayscale → Resize → CLAHE → CNN → Result.
- Accuracy: 98.7 % on a 20 k-note dataset.
- Deploy: Docker container on AWS EC2 + CloudFront CDN.
Morse Code Translator
- Function: Text ↔ Morse conversion in real time.
- Tech: Vanilla JS frontend, Flask backend.
- Extras: audio playback, dark-mode toggle.