Taminul Islam

Taminul Islam

Doctoral Researcher in Computer Science · BASE Lab, Southern Illinois University Carbondale
2026–27 Doctoral Research Fellow
Taminul Islam

I study computer vision and machine learning, with a focus on building lightweight, production-ready systems. My work spans efficient large language and vision-language models, multimodal multi-task learning, and video object segmentation — broadly applicable to real-world problems where compute-efficient perception, multimodal reasoning, and reliable decision-making matter, and I’m open to extending these methods into new domains as the science demands.

I am a PhD student at the School of Computing, Southern Illinois University Carbondale, advised by Dr. Khaled R. Ahmed at the BASE Lab. My research has been published in Nature Scientific Reports, IEEE Transactions on Sustainable Computing, CVPR, WACV, ICCV, IEEE Big Data, and ICMLA, with one article recognized as a Highly Cited Paper for 2024–25. I am an IEEE Graduate Student Member and an ACM Professional Member.

OfficeA307E, Engineering Building, 1230 Lincoln Dr, Carbondale, IL 62901
Emailtaminul.islam [at] siu [dot] edu

At a glance Google Scholar →

36+
Publications
682
Citations
15
h-index
21
i10-index
Last updated: 2026-04-13

Highlights

Oral presentation — CarboFormer at ISVC 2025.
Third place at Falling Walls Lab Illinois — among 16 selected talks.
WACV 2026 — Oral presentation of FUME (HARVEST-Vision Workshop).
Finalist — SIU Three Minute Thesis (3MT) Competition.
Multimodal sensor data collection — livestock and gas-emission research.
Best RSO Award — Bangladesh Student Association at SIUC.
01 / 06

Research Interests

01Computer VisionDetection, segmentation, and recognition with transformer-based architectures.
02Efficient LLMs & VLMsCompressed, low-latency language and vision-language models for edge deployment.
03Multimodal Learning & FusionCross-modal fusion across vision, text, and sensor signals — multi-task heads.
04Video Object SegmentationMemory-efficient transformers for streaming and long-range temporal masks.
05Lightweight Neural ArchitecturesCompact transformer and CNN designs tuned for real-time inference on the edge.
06Sustainable AIQuantifying and reducing the carbon footprint of training and inference.

News All updates →

Apr 2026
New!Paper“TRACE: Thermal Recognition Attentive-Framework for CO₂ Emissions” accepted at V4A @ CVPR 2026.
Apr 2026
New!Paper“FryNet: Dual-Stream Adversarial Fusion for Frying Oil Oxidation” accepted at VISION @ CVPR 2026.
Feb 2026
New!AwardReceived the NSF AIVO Travel Grant ($1,500) to attend WACV 2026.
Feb 2026
New!AwardAwarded the Doctoral Research Fellowship at SIU for the 2026–2027 academic year.
Feb 2026
New!MilestoneCrossed 600+ citations on Google Scholar.
Feb 2026
TalkSelected for the Final of the SIU Three Minute Thesis (3MT) Competition.
Jan 2026
Recognition“Detection and Classification of Cannabis Seeds” recognized as a Highly Cited Paper for 2024–2025.
Jan 2026
Paper“FUME” accepted as oral at HARVEST-Vision @ WACV 2026.
Nov 2025
TalkPresented “CarboFormer” as oral at ISVC 2025, Las Vegas.
Oct 2025
Paper“ArmFormer” accepted at IEEE Big Data 2025.
Sep 2025
AwardReceived Third Place at Falling Walls Lab Illinois among 16 selected talks.
Aug 2025
AwardBest RSO Award — Bangladesh Student Association at SIUC.

Selected Publications All publications →

2026
TRACE: Thermal Recognition Attentive-Framework for CO2 Emissions from Livestock
Taminul Islam, A. Lakhssassi, Toqi Tahamid Sarker, Mohamed Embaby, Khaled R. Ahmed, Amer AbuGhazaleh
V4A @ CVPR 2026
2026
FryNet: Dual-Stream Adversarial Fusion for Non-Destructive Frying Oil Oxidation Assessment
Khaled R. Ahmed, Toqi Tahamid Sarker, Taminul Islam, T. M. Alanezi, Amer AbuGhazaleh
VISION @ CVPR 2026
2026
FUME: Fused Unified Multi-Gas Emission Network for Livestock Rumen Acidosis Detection
Taminul Islam, Toqi Tahamid Sarker, Mohamed Embaby, Khaled R Ahmed, Amer AbuGhazaleh
HARVEST-Vision @ WACV 2026
2025
ArmFormer: Lightweight Transformer Architecture for Real-Time Multi-Class Weapon Segmentation and Classification
Akhila Kambhatla, Taminul Islam, Khaled R. Ahmed
IEEE BigData 2025
2025
CarboFormer: A Lightweight Semantic Segmentation Architecture for Efficient Carbon Dioxide Detection Using Optical Gas Imaging
Taminul Islam, Toqi Tahamid Sarker, Mohamed G Embaby, Khaled R Ahmed, Amer AbuGhazaleh
ISVC 2025
2025
GasTwinFormer: A Hybrid Vision Transformer for Livestock Methane Emission Segmentation and Dietary Classification in Optical Gas Imaging
Toqi Tahamid Sarker, Mohamed Embaby, Taminul Islam, Amer AbuGhazaleh, Khaled R. Ahmed
SEA @ ICCV 2025
2025
WeedSense: Multi-Task Learning for Weed Segmentation, Height Estimation, and Growth Stage Classification
Toqi Tahamid Sarker, Khaled R. Ahmed, Taminul Islam, Cristiana Bernardi Rankrape, Karla Gage
CVPPA @ ICCV 2025
2025
WeedSwin hierarchical vision transformer with SAM-2 for multi-stage weed detection and classification
Taminul Islam, Toqi Tahamid Sarker, Khaled R. Ahmed, Cristiana Bernardi Rankrape, Karla Gage
Scientific Reports (Nature)
2024
Predictive modeling for breast cancer classification in the context of Bangladeshi patients by use of machine learning approach with explainable AI
Taminul Islam, Md. Alif Sheakh, Mst. Sazia Tahosin, Most. Hasna Hena, Shopnil Akash, Yousef A. Bin Jardan, Gezahign FentahunWondmie, Hiba-Allah Nafidi, Mohammed Bourhia
Scientific Reports (Nature)
Read more

Biography

Taminul Islam is a Doctoral Student in Computer Science at the School of Computing, Southern Illinois University Carbondale, working as a Research Assistant in the BASE: Bridging AI, Systems, and Environment lab under the supervision of Dr. Khaled R. Ahmed.

His research journey began during his second year of undergraduate studies, when he became deeply enthusiastic about academic research. Rather than waiting for established opportunities, he founded and led his own student research team — recruiting peers, mentoring junior collaborators, and shaping projects from rough ideas to peer-reviewed publications.

He has collaborated with 25+ faculty members across institutions worldwide, building a reputation for translating ambiguous problems into well-scoped technical contributions. He thrives at the intersection of applied research and engineering — designing systems that hold up under real-world constraints, communicating across disciplinary boundaries with domain experts, and stewarding multi-author projects from inception through peer review. He places a particular emphasis on mentoring: turning curious students into independent researchers and helping them publish their first papers.

Beyond academics, he is passionate about sports and community leadership. He served as Captain of the Bangladesh U-19 National Throwball Team in the 2016 Asian Junior Championship and represented Bangladesh at the 2017 World Games. At SIU, he secured championship victories in Cricket and Badminton, and served as General Secretary of the Bangladesh Student Association — leading the organization to the Best RSO Award in 2025.

After graduation, he plans to join a U.S. National Laboratory as a Research Scientist or pursue a position in academia, contributing to large-scale, mission-oriented research with broad societal impact.