Taminul Islam (he/him)
Taminul Islam is a Doctoral Student in Computer Science at School of Computing, Southern Illinois University Carbondale, currently serving as a Research Assistant at the BASE: Bridging AI, Systems, and Environment Research lab under the supervision of Dr. Khaled R. Ahmed. His research focuses on developing efficient computer vision architectures and transformer-based models for real-world applications. He specializes in semantic segmentation, object detection, multi-task learning, and model optimization—building lightweight, production-ready AI systems that balance accuracy with computational efficiency.
Internship Availability — Summer 2026
I am actively seeking internship opportunities for Summer 2026 in AI/ML/Computer Vision. I'm eager to contribute to cutting-edge projects in deep learning, computer vision systems, and scalable ML deployment. Interested in roles at AI research labs, FAANG research teams, and innovative startups.
Biography
My journey into research began during my second year of undergraduate studies, when I became deeply enthusiastic about academic research. Rather than following traditional paths, I decided to create my own research team from the beginning. Together, we tackled various machine learning and computer vision projects, and I developed a systematic approach: after completing each project, I would convert our work into academic papers and publish them in international conferences. This was how I started my research journey, building a foundation that would shape my entire academic career.
Throughout this journey, I have had the privilege of collaborating with distinguished computer scientists from around the world, including Dr. Khaled R. Ahmed, Professor Azizur Rahman, Dr. Md Zobaer Hasan, Dr. Rejwan Bin Sulaiman, Dr. Inam Ullah Khan, and Prof. Mahbuba Begum. These collaborations have resulted in 31 publications in high-impact venues including Scientific Reports (Nature), IEEE Transactions on Sustainable Computing, ICCV 2025, and leading international conferences, accumulating 490+ citations with an h-index of 14.
My research portfolio focuses on developing advanced computer vision and deep learning solutions that address critical real-world challenges. I have made significant contributions across multiple domains: in medical AI, I developed innovative deep learning architectures for breast cancer classification and early-stage liver disease prediction; in cybersecurity, I worked on fake news detection, social media analysis, and cyber attack prediction for autonomous vehicle perception systems. My recent work at the BASE Lab focuses on building efficient vision transformer architectures, including hierarchical transformers achieving 99.3% accuracy on 200K+ image datasets, lightweight semantic segmentation models with 3x faster inference, and multi-task learning systems. This research has resulted in 2 papers accepted to ICCV 2025 (one Oral presentation, one Poster), demonstrating the impact of efficient model design for production deployment.
Beyond academics, I am passionate about sports and believe in the power of athletic pursuits to build character and leadership. During my undergrad, I achieved a Full Free Scholarship for my sports activities. I actively participate in cricket, soccer, badminton, pickleball, volleyball, and throwball. I served Bangladesh U-19 National Throwball Team (as Captain) in 2016 Asian Junior Championship (Malaysia Tour), and Bangladesh National Throwball Team in 2017 World Games (Nepal Tour). At SIU, I carried forward my sporting legacy—leading my teams to championship victories in Cricket (as Captain) and excelling in Badminton (Doubles) during both the Fall Games 2024 and Spring Games 2025.
Leadership has always been one of my core strengths, whether in sports or academic settings. I served as General Secretary of the Bangladesh Student Association (BSA) at SIUC during the 2024-2025 session, helping our organization achieve the Best RSO Award in 2025 from SIUC. Earlier, I was the Founding General Secretary and Trainer of the Dhaka Collegiate School Debating Club (DCSDC), where I participated in debate competitions alongside my team.
As an active member of the academic community, I serve as Associate Editor for Frontiers in Medicine (Q1 journal), have provided peer review for 15+ manuscripts across top-tier venues, and mentored 20+ students in AI/ML research. I maintain professional membership with ACM and continue to collaborate on cutting-edge research projects. Currently, as a doctoral researcher at Southern Illinois University Carbondale, I focus on advancing computer vision and deep learning techniques, conducting research at the BASE lab that bridges theoretical innovation with practical deployment.
Recent Publications
- Islam, Taminul, Toqi Tahamid Sarker, Khaled R. Ahmed, Cristiana Bernardi Rankrape, and Karla Gage. "WeedSwin hierarchical vision transformer with SAM-2 for multi-stage weed detection and classification." In Scientific Reports 15, no. 1 (2025): 23274.
- Islam, Taminul, Toqi Tahamid Sarker, Mohamed G Embaby, Khaled R Ahmed, and Amer AbuGhazaleh. "CarboNeXT and CarboFormer: Dual Semantic Segmentation Architectures for Detecting and Quantifying Carbon Dioxide Emissions Using Optical Gas Imaging." arXiv preprint arXiv:2506.05360 (2025).
- Syed Mhamudul Hasan, Islam, Taminul, Munshi Saifuzzaman, Khaled R. Ahmed, Chun-Hsi Huang, and Abdur R. Shahid. "Carbon Emission Quantification of Machine Learning: A Review." In IEEE Transactions on Sustainable Computing (2025): 1-19.
- Islam, Taminul, Toqi Tahamid Sarker, Khaled R. Ahmed, Cristiana Bernardi Rankrape, and Karla Gage. "WeedVision: Multi-Stage Growth and Classification of Weeds using DETR and RetinaNet for Precision Agriculture." In 2024 ICMLA, 2024: 1374-1381.
Education
-
Doctor of Philosophy (PhD) in Computer Science, Spring 2024 - Present
Southern Illinois University, Carbondale, USA
-
BSc in Computer Science and Engineering, Spring 2018 - Fall 2022
Daffodil International University, Bangladesh
News
| October 2025 | Our paper "ArmFormer: Lightweight Transformer Architecture for Real-Time MultiClass Weapon Segmentation and Classification" has been accepted in IEEE Big Data 2025. |
| September 2025 | Our paper "CarboFormer: A Lightweight Semantic Segmentation Architecture for Efficient Carbon Dioxide Detection Using Optical Gas Imaging" has been accepted in ISVC 2025. |
| September 2025 | Honored to have been awarded Third Place 🥉 among 16 selected talks in the Falling Walls Lab Illinois competition in Springfield, IL, organized by the Illinois Innovation Network. LinkedIn Reference |
| July 2025 | Two papers have been accepted at ICCV 2025: "GasTwinFormer: A Hybrid Vision Transformer for Livestock Methane Emission Segmentation and Dietary Classification in Optical Gas Imaging", and "WeedSense: Multi-Task Learning for Weed Segmentation, Height Estimation, and Growth Stage Classification". |
| July 2025 | Our article "WeedSwin hierarchical vision transformer with SAM-2 for multi-stage weed detection and classification" has been published in Nature Scientific Reports. Full Text |
| June 2025 | Our collaborative work "Carbon Emission Quantification of Machine Learning: A Review" has been accepted for publication in IEEE Transactions on Sustainable Computing. Full Text |
| June 2025 | New preprint "CarboNeXT and CarboFormer: Dual Semantic Segmentation Architectures for Detecting and Quantifying Carbon Dioxide Emissions Using Optical Gas Imaging" available on arXiv. arXiv |
| June 2025 | Served as an Associate Editor for Frontiers in Medicine (Q1 journal) on a machine learning manuscript for carotid intima-media thickness progression prediction. LinkedIn Reference |
| May 2025 | Secured two CHAMPION trophies for Bangladesh in Cricket and Badminton at the International Spring Games 2025 at Southern Illinois University, Carbondale. Named Most Valuable Player (MVP) in the Badminton event. LinkedIn Reference |
| May 2025 | Presented collaborative research to the Department of Animal Science, SIUC with Toqi Tahamid Sarker and Mohamed Embaby, PhD. Research published in CVPR 2024 and IET Image Processing, bridging School of Computing and School of Agricultural Science. LinkedIn Reference |
| April 2025 | Served as General Secretary of the Bangladesh Student Association (BSA) during 2024-2025 session. BSA received the 'Best RSO Recognition Award' among 17 international RSOs at SIUC for achievements in sports, cultural programs, and food festivals. LinkedIn Reference |