Ultimate Tic-Tac-Toe :
A beginner project to implement a Tic-Tac-Toe game using OpenGL library in C with a recursive type rule addition.
It gave a good hands-on understanding of how graphics based algorithms work. The game logic was quite simple to implement
and selecting the winner just involved an extra step based on recursion. The graphics implementation was a little more
computationaly difficult to implement.
Predicting malignant tumors from Breast Mammograms :
A 2 phase project where we had to first implement preprocessing of the images for feature extraction and tumor mass highlighting.
Followed by a CNN model implementation to predict whether the highlighted mass is malignant, calcination or benign
The preprocessing used Single seeded region growing algorithm to detect pectoral muscles. Multi-level Otsu Thresholding
for object detection and Transfer learning on AlexNet and MobileNet based CNN models for the final predictor.