William C Francis's Projects
An AI-assisted note taking application that does mic calibration, voice to text and AI proofreading
Endless conversations between GPT and Google Gemini with possible human interaction
A fire-fighting robot that scans through its surroundings and extinguishes fire.
Python implementation of Baum Welch algorithm for an HMM
A novel approach to blind motion deblurring by converting a non-blind method (Weiner Deconvolution) to a blind method using Deep Learning
A multi-stage computer vision algorithm to detect edges in images. Developed by John Canny in 1986, widely used in image processing, computer vision, and other fields. This repo provides an implementation in Python
Python program that solves the classic Chess Knight's Tour problem using three different algorithms by also minimizing the time and space complexity.
Pytorch implementation of an all-CNN network trained on the CIFAR-10 dataset to study the nuances of Cosine Annealing technique of learning rate scheduling.
PyTorch implementation of a convolutional deep neural network to colorize grayscale images. Includes sample dataset and code for training, inference, and color temperature control.
Python implementation of an Unscented Kalman Filter (UKF) to track the orientation of a robot in three dimensions given observations from an inertial measurement unit (IMU) that consists of gyroscopes and accelerometers and corresponding data from a motion-capture system called Vicon.
Linear classifier using Support Vector Machines (SVM) which can determine whether an email is Spam or not with an accuracy of 98.7%. Used regularization to prevent over-fitting of data. Pre-processed the E-mails using Porter Stemmer algorithm. Used a spam vocabulary to create a Feature Vector for each E-mail. Prints the top 15 predictors of spam
Pytorch Implementation of various experiments and proposed improvements to the state-of-the-art image super resolution model ESRGAN.
Implementation of Support Vector Machines and Gabor Filters on MNIST dataset
Documentation for the GoKart project
A study of the performance of Gradient Descent, Stochaistic Gradient Descent and Nesterov momentum for a logistic regression model on MNIST dataset
Python implementation of Laplacian pyramid algorithm for blending images using reduce/expand, Gaussian/Laplacian pyramids, and combine/collapse functions for realistic outputs
Modification to the state-of-the-art EfficientPS model by replacing its instance head (Mask-RCNN) with SOLOv2, making it a location based-panoptic segmentation model
Implementation and benchmarking of machine learning methods for identifying hand motions from EEG recordings.
Design of a Neural Network from scratch for recognizing digits written in numpy without using pytorch, keras, or tensorflow libraries
Simultaneous Localization and Mapping (SLAM) in an indoor environment using information from an IMU and a LiDAR sensor collected from a humanoid robot called THOR.
Implementation of the policy iteration algorithm to find the best trajectory for a robot to go from the start to the goal cell in a 2D grid
Python implementation of the Q Learning algorithm with the DDQN trick on OpenAI's version of the cart-pole problem
A Pytorch implementation of a Recurrent Neural Network trained on Leo Tolstoy's War and Peace to imitate the language style used in the book
This project explores the use of a new family of Long Short Term Memory networks applied to the problem of robot localization in the AI Habitat simulator
This is my Pytorch implementation of Segmenting Objects by Locations (SOLO) for instance segmentation
Tic-Tac-Toe game with Single-player and Multi-player modes in Turbo C++
Implementation of Bayes Filter to track the position of a robot operating in a 2D grid world