Khizar Anjum

Hi, I am Khizar and I am a PhD candidate at Rutgers. My interests lie at the intersection of areas of Artificial Intelligence, Analog Neural Networks, Approximate Computing, and Signal Processing.

I currently work at CPS Lab at Rutgers under the supervision of Dr. Dario Pompili. I was awarded an NMF Gold Medal along my Bachelor's from Lahore University of Management Sciences (LUMS) for my outstanding performance.

Email  /  CV  /  GitHub  /  LinkedIn  /  Google Scholar

ResearchGate  /  Projects  /  Blog

🏆 Latest Highlights

Research

I'm interested in computer vision, deep learning, generative AI, and image processing. Most of my research focuses on inferring the physical world from images and signals. My work spans across approximate computing, analog neural networks, and underwater communications.

A novel approach for real-time autonomous navigation using video streams for aerial vehicles.

Creation and evaluation of a comprehensive underwater acoustic communications dataset for research.

A method to identify crowd patterns using drone movement data with minimal computational resources.

An innovative approach for continuous EEG monitoring using battery-less implantable devices.

A framework for enabling aerial robots to make real-time context-aware inferences during operation.

A novel analog neural network architecture for power-efficient cardiovascular health monitoring.

Analysis showing how simple linear models can outperform complex machine learning models for breast cancer diagnosis.

A novel CNN architecture using anisotropic diffusion for efficient EEG signal processing.

A deep learning approach for real-time fault detection and classification in UAVs using FPGAs.

A joint source-channel coding approach for efficient underwater image transmission.

An autoencoder approach for compressing images for underwater acoustic transmission.

A low-power RNN design for health monitoring in wireless environments.

A hybrid approach combining analog and digital sensing for energy-efficient anomaly detection in drones.

A novel neural network architecture using voltage-based resistive processing units and diode activation functions.

Techniques for enabling situational awareness in multi-UAV systems using distributed and approximate computing.

Patents

Recent patents from my research work.

A novel neural network architecture utilizing anisotropic diffusion for efficient computation.

A patented method for efficient image transmission through underwater acoustic channels.

Experience

Professional research experience in academic settings.

Conducting research on embedded systems, neural network optimizations, and underwater communications. Developing novel approaches for battery-less implantable devices and UAV systems.

Worked on machine learning applications for medical image analysis and pattern recognition. Developed algorithms for early detection of neurological disorders using sensor data.

About

I love everything open-source and like to contribute to it a lot. I like how we come together as developers and make amazing things. I like to work on different projects which challenge me in my free-time and like to enhance my skill-set. I have showcased some of them here too. Also, feel free to dig around my blog and do contact me if you have any ideas. Stay strong and stay energetic.

Miscellanea

Academic Service

Teaching