I calculated the camera matrix parameters through utilizing different perspectives of images of checkerboards to obtain extrinsic parameters to calculate intrinsic parameters. I optimized camera parameters by reducing re-projection error by 11%

Successfully created Face Swap in real time through the utilizing state of the art face-warping methods such as Triangulation and Thin Plate Spline

I used the classical Structure From Motion technique to find poses of different views of a camera. I constructed scenes through volumetric representation using images from different angles by implementing NeRF.

I Implemented PB boundary detection algorithm by creating image processing tools: texton, color, and brightness map. I trained and tested a Neural Network in Pytorch to compare sobel and canny edge detection algorithm.