Greetings 🖖. I am a recent graduate with a Bachelor's degree in Computer Systems Engineering. I am a dedicated researcher with a focus on the intersection of security and machine learning. I have conducted extensive research at accredited labs and am now eager to further my academic journey in a graduate program.
I conducted research focused on energy-harvesting wireless sensor networks (EHWSNs). In the course of this project, I developed and assessed node energy control methods utilizing discrete-event simulation, specifically OMNET++, incorporating routing functionality. The research involved executing simulation experiments, documenting designs, and analyzing results to enhance energy management strategies in EHWSNs.
I participated in two projects, namely "Smart Lock" and "Anomaly Detection in IIoT Systems." In the later project, I conducted research on Industrial IoT detection models, exploring rule-based, unsupervised, and supervised approaches while assessing their respective strengths and weaknesses. Through this research, I identified specialized federated device models that exhibited superior accuracy, low false positives, and compatibility with edge computing environments.
Proposed and implemented a federated learning approach to address data privacy concerns in medical imaging for Diabetic Retinopathy. Implemented Differential Privacy (DP) with the addition of Gaussian noise to measure privacy leakage and evaluate privacy budget consumption.
Developed real time detector which can detect if people are wearning PPE (Personal Protective Equipment) in a pharmaceutical company. Created an efficient YOLOv4 computer vision model for real-time object detection. Applied annotation and augmentation to the entire training set and converted detector weights to TensorFlow format. Integrated live detection features, including real-time object counting and record-keeping.
Lead the design and implementation of advanced computer vision algorithms and models to address complex challenges in image and video analysis.Prepare and clean large datasets for model training, conducting thorough exploratory data analysis (EDA) as needed.
I contributed by implementing "Endpoint Detection and Response" solution using open-source tools as a practical alternative to commercial EDR solution, to reduce investigation times and improve endpoint threat detection. I have implemented EDR by using 3 open source technologies. Osquery; to query a database of recorded systems, Fleet; to manage Osquery through the creation of queries and packs, and Redis; to store, search, and analyze big volumes of data quickly and in near real time.
CGPA : 3.68 / 4.00
Diabetic Retinopathy Diagnostics using Federated Learning Approach
Bushra Ashfaque, Dr. Majida Kazmi, Dr. Saad Ahmed Qazi