Azizul Zahid

I am a researcher in Human Computer Interaction (HCI) and AI at EPIC Lab, Electrical Engineering and Computer science (EECS), University of Tennessee Knoxville (UTK).

I am a fourth-year PhD student in Computer Engineering working with Dr. Sai Swaminathan. I also received my Master’s degree in Computer Engineering from UTK in Summer 2025.

My reserach area lies in on device LLM inference for task specific Human-AI Interaction, focusing on Intelligent guidance systems. I have substantial experience on implementation of LLM on resource constraint devices to build intelligent werable and tiny systems. Additionally, I have built an end-to-end intelligent robots prototyping with utilizing Reinforcement Learning (RL) algorithms.

I earned my B.Sc. from Bangladesh University of Engineering and Technology (BUET) majoring Electrical and Electronic Engineering (EEE).

📢 I am actively looking for internship opportunities. If you know any of opportunities, feel free to reach out to me.

news

Aug 16, 2025	Received my M.Sc. in Computer Engineering.
Jun 03, 2025	Received NSF Travel Award for CHASE’25 conference.
Apr 22, 2025	My workshop paper got accepted to ICRA 25 conference.
Mar 10, 2025	My first phd paper as a lead author got accepted to CHASE 25 conference.
Feb 01, 2025	Submitted another paper on RadioGami system to IMWUT’25.

Selected publications

ArXiv’26 preprint

ProcAgent: An Agentic Framework for Procedural Task Guidance on Edge with Human-in-the-Loop

Azizul Zahid, Subrata Biswas, Bashima Islam, and 1 more author

arXiv preprint arXiv:2607.24770, 2026

Bib PDF

@article{zahid2026procagent,
  title = {ProcAgent: An Agentic Framework for Procedural Task Guidance on Edge with Human-in-the-Loop},
  author = {Zahid, Azizul and Biswas, Subrata and Islam, Bashima and Swaminathan, Sai},
  journal = {arXiv preprint arXiv:2607.24770},
  year = {2026},
}

ICRA’25

Toward Aligning Human and Robot Actions via Multi-Modal Demonstration Learning

Azizul Zahid, Jie Fan, Farong Wang, and 3 more authors

In ICRA 2025 Workshop: Human-Centered Robot Learning in the Era of Big Data and Large Models, 2025

Bib PDF

@inproceedings{zahid2025toward,
  title = {Toward Aligning Human and Robot Actions via Multi-Modal Demonstration Learning},
  author = {Zahid, Azizul and Fan, Jie and Wang, Farong and Dy, Ashton and Swaminathan, Sai and Liu, Fei},
  booktitle = {ICRA 2025 Workshop: Human-Centered Robot Learning in the Era of Big Data and Large Models},
  year = {2025},
  url = {https://openreview.net/forum?id=ZyRVRYAC97},
}

CHASE’25

Pulseride: A robotic wheelchair for personalized exertion control with human-in-the-loop reinforcement learning

Azizul Zahid, Bibek Poudel, Danny Scott, and 4 more authors

In Proceedings of the ACM/IEEE International Conference on Connected Health: Applications, Systems and Engineering Technologies, 2025

Bib PDF

@inproceedings{zahid2025pulseride,
  title = {Pulseride: A robotic wheelchair for personalized exertion control with human-in-the-loop reinforcement learning},
  author = {Zahid, Azizul and Poudel, Bibek and Scott, Danny and Scott, Jason and Crouter, Scott and Li, Weizi and Swaminathan, Sai},
  booktitle = {Proceedings of the ACM/IEEE International Conference on Connected Health: Applications, Systems and Engineering Technologies},
  pages = {57--68},
  year = {2025},
}

IMWUT’24
NeuroCamTags: Long-Range, Battery-free, Wireless Sensing with Neuromorphic Cameras

Danny Scott, Matthew Bringle, Imran Fahad, and 3 more authors

Proc. ACM Interact. Mob. Wearable Ubiquitous Technol., Sep 2024

Abs DOI Bib PDF

In this research, we introduce NeuroCamTags, a battery-free platform designed to detect a range of rich human interactions and activities in entire rooms and floors without the need for batteries. The NeuroCamTag system comprises low-cost tags that harvest ambient light energy and utilize high-frequency modulation of light-emitting diodes (LEDs) for wireless communication. These visual signals are captured by an available neuromorphic camera, which boasts temporal resolution and frame rates an order of magnitude higher than those of conventional cameras. We present an event processing pipeline that allows simultaneous localization and identification of multiple unique tags. NeuroCamTags offer a wide range of functionalities, providing battery-free wireless sensing for various physical stimuli, including changes in temperature, contact, button presses, key presses, and even sound cues. Our empirical evaluations demonstrate impressive accuracy at long ranges up to 200 feet. In addition to these findings, we consider a range of applications such as battery-free input devices, tracking of human movement, and long-range detection of human activities in various environments such as kitchens, workshops, etc. By reducing reliance on batteries, NeuroCamTags promotes eco-friendliness and opens doors to exciting possibilities in smart environment technology.
@article{IMWUT24_Scott, author = {Scott, Danny and Bringle, Matthew and Fahad, Imran and Morales, Gaddiel and Zahid, Azizul and Swaminathan, Sai}, title = {NeuroCamTags: Long-Range, Battery-free, Wireless Sensing with Neuromorphic Cameras}, year = {2024}, issue_date = {September 2024}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, volume = {8}, number = {3}, url = {https://doi.org/10.1145/3678529}, doi = {10.1145/3678529}, journal = {Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.}, month = sep, articleno = {122}, numpages = {25}, keywords = {Activity sensing, Battery-free, Context Aware Computing, Wireless sensing} }