Titon Barua
Who am I?
Researcher, hacker, problem-solver, polyglot programmer. I love exploring intersections of different disciplines — ML and deep learning, heterogeneous and embedded systems programming, compilers, optimization problems, graph algorithms, computer vision and robotics.
Back in 2010, I started my career as a PHP programmer while getting a bachelor's degree in mechanical engineering. Ever since, my curiosity and fascination with computing technology and computer science has only grown stronger. After working as a software developer in various domains and technologies for close to a decade, I found functional programming to be the most sustainable approach to tackle complexity. These days, I prefer to program in Clojure for general computing, Python for numerical analysis and machine learning, and C/C++ for close-to-metal programming.
In 2022, I started my graduate studies in computer science, largely inspired by my first peer reviewed scientific work in 2020. As of 2026, I am a doctoral candidate researching data-flow compilation techniques for specialized neural network accelerator hardware with a focus on low-latency, power-efficient inference.
Education
- PhD Candidate - UofSC (as of 2026)
- Master in Computer Science - UofSC (2024)
- Post-Baccalaureate Certificate in AI - UofSC (2024)
- Bachelor in Mechanical Engineering - BUET (2013)
Work Experience
- Graduate Research Assistant
Heterogeneous and Reconfigurable Computing Lab, Molinaroli College of Engineering and Computing, University of South Carolina
Advisor: Prof. Dr. Jason D. Bakos
Domain: Hardware accelerated deep learning, ML data-flow compilers, Graph Optimization, Code generation, DSP programming
Tech: C++, Clojure, ONNX, Pytorch, Vitis Tools, AMD AI Engine
Dec 2024–Current - Graduate Research Assistant
Autonomous Field Robotics Lab, Molinaroli College of Engineering and Computing, University of South Carolina
Advisor: Dr. Ioannis Rekleitis
Domain: Computer vision, Structure from Motion, Underwater robotics, Mechanical and electronic design and fabrication
Tech: C/C++, Python, ROS, eigen, numpy, clifford(geometric algebra library), micropython, OpenSCAD
Aug 2022–Nov 2024 - Developer Lead / Founder
M/s Hammock Lab, BD / Hammock Lab Ltd, UK (www.dandydialer.com)
Domain: IP Telephony, Realtime communication, Scalable backend architecture
Tech: Clojure, Clojurescript, Datomic, Linux, Freeswitch, LXD
2019–Jul 2022 -
System Integration Consultant / CTO
M/s Concitus, Dhaka, BDDomain: HPC, ML and embedded systems programming, Digital Design
Tech: C, Python, GNU Scientific Library, Numpy, Raspberry Pi, Micro-controllers, SPI, UART, I2C, Point-Of-Sale Terminals
2017–2019 -
Full-stack Developer
Vimmaniac Consultancy ( vimmaniac.com )Domain: Web and internet application development
Tech: Linux, PHP, Python + flask, Javascript, MySQL, PostgreSQL
2010–2016
Publications
-
A High Throughput and Unbiased Machine Learning Approach for
Classification of Graphene Dispersions, 2020
Md. Joynul Abedin, Titon Barua, Mahdokht Shaibani, Mainak Majumder
doi.org/10.1002/advs.202001600 -
Customized Production of Holey Graphene Oxides via a Continuous Flow Process, 2023
Wanqing Chen, Md. Joynul Abedin, Titon Barua, Meysam Sharifzadeh Mirshekarloo, Sally El Meragawi, Mainak Majumder
doi.org/10.1002/smll.202304227 -
Weakly Supervised Caveline Detection For AUV Navigation Inside Underwater Caves, 2023
Boxiao Yu, Reagan Tibbetts, Titon Barua, Ailani Morales, Ioannis Rekleitis, Md Jahidul Islam
doi.org/10.48550/arXiv.2303.03670 -
Caveline Detection at the Edge for Autonomous Underwater Cave Exploration and Mapping, 2023
Mohammadreza Mohammadi, Sheng-En Huang, Titon Barua, Ioannis Rekleitis, Md Jahidul Islam, Ramtin Zand
doi.org/10.1109/ICMLA58977.2023.00210 -
CaveSeg: Deep Semantic Segmentation and Scene Parsing for Autonomous Underwater Cave Exploration, 2024
Adnan Abdullah; Titon Barua; Reagan Tibbetts; Zijie Chen; Md Jahidul Islam; Ioannis Rekleitis
doi.org/10.1109/ICRA57147.2024.10611543