Here are a selection of projects that I have worked on over the years.
This project uses a custom-trained YOLOv8 model to detect and track rats in video streams, compute the distance traveled, and generate visual analytics such as path plots and heatmaps.
This code implements a sophisticated signal processing system on an STM32 microcontroller, designed to optimize signal reception and communication through advanced combining techniques.
A machine learning project utilizing a hybrid CNN-LSTM architecture with an attention mechanism for explainable AI.
The School Management System is a Django-based web application designed to streamline day-to-day academic administration. The project includes a simple, responsive static front-end, built with plain HTML, CSS and JavaScript, served from Django’s static/ directory.
Design and implementation of a VLC communication module to transmit data using the STM32 microcontroller platform.
This code is a script for training an image classification model MobileNet using PyTorch. The code is structured to facilitate easy training, evaluation, and monitoring of a deep learning model for image classification.
This comprehensive code implements a deep learning solution for spoofing detection using a Vision Transformer (ViT) architecture. The system leverages transfer learning by fine-tuning a pre-trained ViT model for binary classification, distinguishing between genuine and spoofed images in a specialized security dataset.
This documents the training and evaluation of a Hybrid CNN-LSTM Attention model for time series classification in a dataset. The model combines convolutional neural networks (CNNs) for feature extraction, long short-term memory (LSTM) networks for sequential modeling, and attention mechanisms to focus on important parts of the sequence.
Fundamental files to train and evaluate a simple LSTM, MLP, CNN, and RNN model which can be trained on a time-series dataset composed of n input features and m outputs classes.
You can browse all of my source code, ongoing work, and raw scripts directly on my GitHub profile.