Computer Science Projects

Implemented the 'Procedural Noise using Sparse Gabor Convolution' by A. Lagae and Al. introducing a new procedural noise.

Created all of the user interface and generated the animation starting from a woman 3D model to create a dance animation loop. It also included some cloth physically based animation.

Implemented the 'Variational Shape Approximation' by D. Cohen-Steiner, P. Alliez & M. Desbrun, introducing a new way to approximate shapes.

A rendering engine programmed from scratch in C++ using raytracing method with the Phong shading model, in addition to refraction and reflection rays. The simple geometries (sphere, planes, triangle, etc..) are computer generated. (Source files: Copyright (C) 2022 by Hadi Fadaifard)

Physical model of a cloth with spring forces and computation of the contacts with spheres and cylinders.

Computed the skinning around different skeletons according to the movement using Linear Blend Skinning.

Refocus system using different focus images, measuring focus using modified Laplacian filter and choosing best focus image in Matlab. (Image source: https://www.cs.columbia.edu/CAVE/projects/focalsweep/)

Automatic generation of 3D shapes on pre-filmed or live filmed surfaces, using AR pipelines on OpenCV in C++.

Procedural generation of a forest terrain and animal animations in C++ using OpenGL and VCL.

Generated 3D hair shapes using Bezier Curves on the normal circular sections.

Subdivided a spherical mesh using the loop subdivision as well as an adapted method for spheres.

Line detection using Hough transform in Matlab. (Image source: https://boardgamegeek.com/image/113720/checkers)

Generated randomized falling spheres and simulated their collisions with other spheres and with the box walls around using a physically based model.

Matching the features of two pictures to create a panorama applying RANSAC homography.

Point alignment using Iterative Closest Point (ICP) and normal estimation using Principal Component Analysis (PCA).

Implemented a canny edge detector to draw the edges of a given picture.

Implemented an interactive FFD for several different 3D geometric shapes.

Motion tracking using color histograms matches in Matlab. (Source video: https://www.youtube.com/watch?v=e_WBuBqS9h8&t=1568s)

Programmed linear blend shapes to move from one facial expression to another.

Offered a solution to the Graph Drawing Challenge 2021, aimed at finding the optimal graph drawing to minimize a certain graph metric.

Visualization of the cancer data available on the GDC Portal using javascript and the D3 module.

Implemented local deformers such as translating, twisting and scaling, orthogonally to the view or the geometry.

Physically based simulation of bouncing spheres.

Kaggle competition to predict the number of retweets of a collection of covid-related tweets using Python.

Optimized performance of two different models: the VGGNet for visual recognition, and the LSTM for sentiment analysis.

Built and compared different models to predict if a song will be popular.

Implemented a website to do grocery shopping online using our own designed database. The project required some SQL queries, and some Python/HTML programming to code a Flask app.

Designed a mobile application from scratch. The app was designed to help dancers create their choreographies. We mainly used Figma for the UI design, and we started a quick working prototype using JavaScript.