S.M. Tawhid
Physics & Bio-Inspired Intelligent Systems · Autonomous Multi-Agent Systems · Avian Collision Avoidance
smtawhid77megh@gmail.com | Google Scholar | ORCID | GitHub | LinkedIn | ResearchGate | +8801709006947 | Dhaka, Bangladesh
Profile
Researcher in Physics and Bio-Inspired Intelligent Systems, with a central focus on autonomous multi-agent systems and avian collision-avoidance behavior analysis. Current work bridges stereo vision, 3D trajectory reconstruction, and behavioral modeling to uncover the kinematic principles birds use to navigate dense flocks — and to translate those principles into bio-inspired swarm MAV control strategies. Broader expertise spans reinforcement learning, explainable AI, federated learning for medical imaging, and computer vision (YOLO & OpenCV). Passionate about applying machine learning to build intelligent, scalable, and real-world autonomous systems grounded in biological insight.
Experience
- Directed organizational strategy, research initiatives, and institutional growth of NeuroFlight Lab.
- Led and mentored 10+ multidisciplinary research teams, contributing to multiple submissions in Scopus-indexed journals.
- Supervised software development operations, including the AIUB Journal of Science and Engineering (AJSE, Q3) platform and several AI-driven software solutions.
- Organized international seminars featuring global industry leaders and academic experts, enhancing international collaboration and research visibility.
- Advanced autonomous flight systems research through reinforcement learning and real-time adaptive UAV control for stability and navigation.
- Supported editorial operations and security management for the AIUB Journal of Science and Engineering (AJSE, Q3).
- Contributed to the development of IoT-based automation systems, including smart chair technologies.
- Conducted bird flight research using computer vision, motion tracking, and trajectory analysis methodologies.
- Developed a personalized stress detection framework by integrating facial micro-expression dynamics with physiological signals, including ECG, EDA, and respiration data.
- Implemented an interpretable multimodal deep learning architecture combining LSTM and dense neural networks for robust stress classification.
- Collected, processed, and analyzed data from the WESAD dataset and custom webcam recordings of healthcare professionals to enhance real-world applicability.
- Conducted research on few-shot learning methodologies for retinal vessel segmentation under limited-data conditions.
- Implemented, optimized, and evaluated U-Net architectures and advanced variants on medical imaging datasets.
- Developed efficient segmentation frameworks tailored for low-sample clinical environments to improve diagnostic applicability.
- Participated in technical workshops, professional seminars, and collaborative initiatives focused on engineering skill development, leadership, and professional networking.
- Contributed to community-driven engineering programs, knowledge-sharing sessions, and project exhibitions promoting innovation, collaboration, and technical excellence.
Research & Publications
-
A stereo dataset of annotated budgerigar flight trajectories for multi-agent collision avoidance studies
Q1Published
Data in Brief — Elsevier
-
Multi-agent collision avoidance kinematics in budgerigars: toward bioinspired swarm MAV control
Q1Published
Bioinspiration & Biomimetics — IOP Publishing
-
Phys-Reg Framework for Biomimetic Bird Flight Forecasting
Q1Under Review
Results in Engineering — Elsevier
-
Do birds in larger flocks turn less but turn together? Group-size-dependent trade-offs between individual maneuvering and collective turning in budgerigars
Q2Under Review
Behaviour — Brill Publishers
-
Extracting Effective Stability Landscapes from Fish Schools: Diagnostic Insights for Tunable Collective Behaviour in Bio-Inspired Robotic Swarms
Q1Under Review
Bioinspiration & Biomimetics — IOP Publishing
-
TumorRes-UNet: Explainable Residual U-Net for Robust Multimodal Brain Tumor Segmentation in MRI
Q2Under Review
Global Health Journal — Elsevier
-
Feature Fusion Hybrid Deep Learning Model for Brain Tumor Classification from MRI Scans
Q1Under Review
IEEE Access
-
Explainable Hybrid CNN-Transformer Framework for Papaya Leaf Disease Classification with Layer-Wise Grad-CAM Analysis
Q3Published
AJSE
-
Xception-Based Brain Tumor Classification with GAN Augmentation and Multi-Technique Explainable AI
Q3Accepted
AJSE
-
Multi-Class Infectious Skin Disease Classification Using Transfer Learning and Explainable Artificial Intelligence
Q3Submitted
Healthcare Engineering Letters
-
Video-based Flight Pattern Mining of Birds in Controlled Tunnel Environments Using YOLOv8 and Data Analytics
Submitted
4th International Conference on Computing Advancements (ICCA)
-
Explainable Machine Learning for Academic Risk Prediction Using Leakage-Aware Grade Trajectory Features
Submitted
4th International Conference on Computing Advancements (ICCA)
-
Personalized Stress and Anxiety Detection Using Temporal Micro-Expression Dynamics with Explainable AI
Accepted & Presented
INTI International University, Malaysia
Technical Projects
Real-time face detection system from live webcam video streams.
LDA-based topic modeling on recent news articles to uncover hidden themes.
Café management system with splash screen, role-based login (Admin/User/Guest), and database integration.
Dynamic web application for event registration, listings, and admin control.
Education
- B.Sc. (Eng.), Computer Science and Engineering
American International University-Bangladesh (AIUB), Dhaka
2021 – 2026 • CGPA 90%
Major: Information System (Data Science) • Minor: Computer Vision and Pattern Recognition - Higher Secondary School Certificate (Science)
Engineering University School and College, BUET Campus, Dhaka
2018 – 2020 • GPA 5.00/5.00 - Secondary School Certificate (Science)
Akkelpur F.U. Govt. Pilot School, Joypurhat
2016 – 2018 • GPA 5.00/5.00 - Junior School Certificate (Science)
Akkelpur F.U. Govt. Pilot School, Joypurhat
2013 – 2015 • GPA 5.00/5.00 • JSC Scholarship (Govt. of Bangladesh) - Primary School Certificate (Science)
Sri Kristopur Govt. Primary School, Joypurhat
2007 – 2012 • GPA 5.00/5.00 • PSC Scholarship (Govt. of Bangladesh)
Skills & Interests
Coding Skills
Research & Development
Frameworks & Libraries
Tools & Typesetting
Languages
Interests
Honors & Awards
- Best Poster Award – Thesis Day (Summer 2024–2025) — awarded for the best thesis among 100+ undergraduate groups.
- Government Scholarship (JSC) – awarded for outstanding academic performance, Government of Bangladesh.
- Government Scholarship (PSC) – awarded for outstanding academic performance, Government of Bangladesh.
Research Appointments
- Undergraduate Researcher
American International University-Bangladesh (AIUB), Dhaka
Supervised by Dr. Debajyoti Karmaker — 2 years - Undergraduate Researcher
American International University-Bangladesh (AIUB), Dhaka
Supervised by Prof. Dr. Dip Nandi — 2 years - Undergraduate Researcher
American International University-Bangladesh (AIUB), Dhaka
Supervised by Mahfujur Rahman — 1 year
References
- Dr. Debajyoti Karmaker
Sessional Academic, School of Computing Technologies, RMIT University, Australia
Sessional Lecturer, Australian Institute of Higher Education, Australia
Former Associate Professor & Head, Undergraduate Program, FST, AIUB
d.karmaker@aiub.edu - Prof. Dr. Dip Nandi
Professor & Dean, Faculty of Science and Technology, AIUB
dip.nandi@aiub.edu - Mahfujur Rahman
Assistant Professor, AIUB
mahfuj@aiub.edu