Projects

Developed a Brain-Computer Interface (BCI) system using EEG data to identify user-intended characters using P300 event-related potentials. Used BCI Competition III dataset for training and testing.

Implemented lane detection using Hough Transform after enhancing visibility in low-light images via Retinex-based image enhancement techniques.

Created an adaptive Otsu binarization algorithm by tiling and analyzed effects of Gaussian and Bilateral filtering on noisy images.

Fine-tuned pretrained convolutional neural networks on a custom dataset to perform accurate image classification using Keras and TensorFlow.

Designed a real-time air pollution monitoring system using Arduino and gas/humidity sensors, and displayed data on a live web dashboard.

Simulated and implemented a DSB-C amplitude modulator using a Gilbert Cell multiplier and an envelope detector. Tested on breadboard after NI Multisim simulation.

Projects include MNIST digit classification, COVID-19 case prediction, transferred learning for dog breed classification, and Twitter sentiment analysis.

Used OpenCV and facial landmark detection to create a version of the Chrome Dino game controlled by eye-blinks. Useful as an accessibility demo.

Rebuilt the nostalgic Snake game using Python and OOP concepts. A testbed for future reinforcement learning experiments.