Hello, I'm Obaidul Haque
AI Engineer and Researcher
Medical Imaging & Computer Vision · Natural Language Processing · Biomedical Signals
Dedicated to conducting research and contributing to AI-related projects across Medical Imaging & Computer Vision, Natural Language Processing, and Biomedical Signal Processing.
Current Focus & Research Pillars
Conducting peer-reviewed deep learning research and contributing to AI-related projects across Medical Imaging & Computer Vision, Natural Language Processing, and Biomedical Signals.
I am an early-career AI Engineer and Researcher focused on conducting deep learning research and contributing to impactful AI-related projects. I completed my B.Sc. in Computer Science & Engineering from IUBAT with a CGPA of 3.89/4.00 across 144 credit hours.
My primary research focus spans Medical Imaging and Computer Vision—specifically dose-robust Low-Dose CT reconstruction using state-space Mamba models, brain MRI Alzheimer's staging, and cellular classification—alongside Biomedical Signal Processing featuring dual-lead CNN-Transformer ECG arrhythmia detection with gradient-based explainability.
In Natural Language Processing (NLP), I explore transformer representations, BanglaBERT linguistic modeling, and semantic sentence-pair analysis. Alongside research, I actively build full-stack systems with Java Spring Boot and React, and maintain sharp problem-solving habits with 550+ problems solved across Codeforces, LeetCode, and ICPC preliminaries.
Department of Computer Science & Engineering
Medical Imaging & Computer Vision
CT Dose & VisionPioneering architectures like OBAIS-Mamba for dose-robust low-dose CT reconstruction, brain MRI Alzheimer disease staging, and fine-grained cellular classification.
Natural Language Processing (NLP)
Transformers & LLMsInvestigating transformer architectures, BanglaBERT representations, aspect-based sentiment and abusive text detection, and semantic sentence-pair modeling.
Biomedical Signal Processing
ECG & Bio-SignalsFormulating interpretable dual-lead CNN-Transformer hybrids for ECG arrhythmia classification with Grad-CAM explainability, stroke risk prediction, and EEG neural dynamics.
Highlighted Research Publications
Neural architectures spanning state-space continuous modeling, clinical epidemiology, and interpretable biomedical signals.
OBAIS-Mamba: Kernel-Invariant Anatomical State Learning for Dose-Robust Low-Dose CT Reconstruction
ACM International Conference on Computing Advancements (ICCA), 2026
Proposes a novel state-space model (SSM) architecture, OBAIS-Mamba, engineered to overcome quantum noise and severe streak artifacts in Low-Dose Computed Tomography (LDCT). By formulating a kernel-invariant continuous state-space formulation, the model learns anatomical continuity across slices while preserving delicate lesion edges without over-smoothing.
Stroke Prediction Using Ensemble Machine and Deep Learning Models
Biomedical Materials & Devices (BMD), Springer Nature, 2025
Comprehensive epidemiological and clinical study benchmarking stacking ensemble classifiers, gradient boosting algorithms, and deep neural networks to predict cerebrovascular stroke risk factors across heterogeneous demographic cohorts.
An Interpretable Dual-Lead CNN-Transformer Hybrid Framework for ECG Arrhythmia Classification with Gradient-Based Explainability
IEEE SPICSCON, 2026
Presents a hybrid architecture fusing convolutional local receptive fields with multi-head self-attention transformers for dual-lead electrocardiogram (ECG) arrhythmia categorization. Incorporates gradient-weighted class activation mapping (Grad-CAM) to visualize P-Q-R-S-T morphological anomalies.
Featured Systems & Applications
Production full-stack systems combining Java Spring Boot backends, local LLM integrations, and reactive web architectures.
EduMaster
AI-Assisted Learning Management System
Production-grade full-stack LMS developed with Spring Boot, React, and PostgreSQL, showcased and defended before the IUBAT Practicum Board in January 2026. Integrates local LLM intelligence via LangChain4j and Ollama for automated context-aware learning assistance, course Q&A, and personalized career roadmaps.
Research Tracker
Academic Research Management & Analytics Platform
Cross-platform desktop & web platform for managing academic research lifecycles. Automates metadata ingestion, co-author intelligence, and paper tracking with DOI-based OpenAlex integration.
Tools, Models and Technologies
The deep learning frameworks, transformer architectures, programming languages, and backend engines I work with.
PyTorch
2+ yearsDeep Learning
Mamba
SpecializedSelective State Space
State Space Models
SpecializedSequence Modeling
Hugging Face
2+ yearsTransformers Hub
BanglaBERT
SpecializedBengali Linguistic AI
OpenCV
2+ yearsComputer Vision
Torchvision
2+ yearsVision Datasets
SciPy
SpecializedDigital Signal DSP
Grad-CAM
SpecializedModel Explainability
LangChain4j
ProductionLLM Orchestration
Ollama
ProductionLocal LLM Serving
TensorFlow
2+ yearsDeep Learning
Keras
2+ yearsNeural Networks API
Scikit-Learn
2+ yearsMachine Learning
NumPy
3+ yearsVectorized Mathematics
Pandas
3+ yearsData Analysis
Explore More Sections
Deep dive into specific milestones, publications, and background details.
Education & Curriculum
IUBAT B.Sc. in CSE with CGPA 3.89 / 4.00, 144 credit hours completed, and verified transcript grades.
Work Experience
Full Stack AI Developer at Saask Task (Canada Remote) and Frontend Intern at AKS Global IT Services.
Research & Publications
40+ scientific papers across ACM, IEEE, and Springer Nature in Medical Imaging, NLP, and Bio-Signals.
Competitive Programming & Leadership
550+ problems solved across Codeforces and LeetCode, ICPC Preliminary, and BASIS IUBAT Executive Member.
Hobbies & Interests
Lo-Fi acoustic soundscapes, street geometry photography, chess tactics, and reading AI research preprints.
Contact & Collaboration
Direct email dispatch, one-click copy, verified WhatsApp line, and professional social networks.