PhD Researcher · CRIL, Université d'Artois, France

Md Yasir Arafat

PhD Researcher in Graph Neural Networks & 3D Mesh Learning

CRIL – Centre de Recherche en Informatique de Lens, Université d'Artois, France  ·  Assistant Professor, JUST, Bangladesh

Working at the intersection of Graph Neural Networks, 3D Mesh-Based Learning, and Scientific Machine Learning. My PhD explores GNNs for non-uniform 3D objects — including mesh pooling strategies, graph construction from 3D meshes, and temporal GNN applications in fluid simulation, weather modelling, and physical simulation.

Graph Neural Networks 3D Mesh Learning Deep Learning Scientific ML Computer Vision AI for Engineering
PhD at CRIL → Research Publications Contact
7
Publications
2024
PhD Start Year
2
Research Grants
6+
Years Teaching
Md Yasir Arafat

About

Academic Background

Md Yasir Arafat is a PhD Researcher at the Centre de Recherche en Informatique de Lens (CRIL), Université d'Artois, France, where he began his doctoral studies in 2024. His thesis — "Apprentissage sur des objets 3D non-uniformes : une méthode basée sur les graphes" (Learning on Non-Uniform 3D Objects: A Graph-Based Method) — is supervised by Prof. Saïd Jabbour and co-supervised by Dr. Wissem Inoubli, funded by Université d'Artois.

His PhD research focuses on Graph Neural Networks (GNNs) for modelling non-uniform 3D objects such as 3D meshes, addressing scalability challenges through mesh-specific pooling strategies and graph construction methods. The work extends to temporal GNN applications in fluid simulation, weather modelling, and physical 3D mesh simulation.

In parallel, he holds the position of Assistant Professor in the Department of Computer Science and Engineering at Jashore University of Science and Technology (JUST), Bangladesh. He teaches Artificial Intelligence, Machine Learning, Pattern Recognition, Computer Vision, and Bioinformatics, and has supervised numerous undergraduate and Master's-level research projects.

He has undertaken advanced training at IIT Kharagpur and IIT Guwahati through the Bangladesh–Bharat Digital Service and Employment Training (BDSET) programme, working on ML/DL/AI and extended reality (XR) technologies. His internship under Prof. Pabitra Mitra focused on cross-lingual text summarization for Bangla–English language pairs.

Education

2024 – Present
PhD in Computer Science Current
CRIL, Université d'Artois, France
Supervisor: Saïd Jabbour  ·  Co-supervisor: Wissem Inoubli  ·  Funding: Artois
Thesis: Learning on Non-Uniform 3D Objects: A Graph-Based Method
2013 – 2015
M.Sc. in Computer Science and Engineering
Jashore University of Science and Technology
CGPA: 3.82 / 4.00  ·  Position: 2nd
Thesis: Forecasting Civil Unrest Events with Future Time by Analyzing Social Media
2009 – 2012
B.Sc. in Computer Science and Engineering
Jashore University of Science and Technology
CGPA: 3.81 / 4.00  ·  Position: 2nd with Distinction
Thesis: Lossless Efficient Data Compression Techniques

Quick Info

StatusPhD Researcher
LabCRIL, Artois
PhD Start2024
SupervisorSaïd Jabbour
Co-supervisorWissem Inoubli
FundingUniversité d'Artois
AlsoAsst. Prof., JUST
NationalityBangladeshi

Research Interests

Machine Learning Deep Learning Graph Neural Networks Graph Representation Recommendation Systems Computer Vision Scientific ML AI for Engineering

Connect

Research

Research Areas

Exploring the frontiers of machine learning applied to complex scientific, engineering, and real-world problems.

PhD Thesis · CRIL, Université d'Artois · 2024 – Present

Apprentissage sur des objets 3D non-uniformes : une méthode basée sur les graphes

"Learning on Non-Uniform 3D Objects: A Graph-Based Method"

This thesis explores GNNs for modelling non-uniform 3D objects such as 3D meshes. It addresses scalability challenges through mesh-specific pooling strategies and graph construction methods, and extends to temporal GNN applications in fluid simulation, weather modelling, and physical 3D mesh simulation.

Supervisor: Saïd Jabbour Co-supervisor: Wissem Inoubli Funding: Université d'Artois Axis: Data Learning & Processing
MeshGraphNets 3D Mesh Pooling Graph Construction Temporal GNNs Fluid Simulation Physical Simulation Weather Modelling Medical 3D Images
View at CRIL →   ADUM Listing →

Graph Neural Networks

Core PhD research area. Advancing GNNs for non-uniform 3D mesh data, with mesh-specific pooling strategies and scalable graph construction methods that surpass CNN limitations on irregular geometries.

MeshGraphNets Graph Transformers Message Passing Geometric Deep Learning Mesh Pooling

Scientific Machine Learning

Developing AI-driven surrogate models and physics-aware learning systems that accelerate scientific simulation and encode physical principles within neural architectures.

Physics-Informed Learning Surrogate Modeling PDE Learning Digital Twins Engineering Simulation

Computer Vision

Applying convolutional and transformer-based vision models to real-world classification, detection, and segmentation problems, including agricultural and biomedical image analysis.

CNN Architectures Image Classification Object Detection YOLOv8 Segmentation

Deep Learning

Researching neural network architectures, transformer models, and self-supervised learning strategies for language understanding, sequence modeling, and structured prediction.

Transformers RNNs / LSTMs Self-Supervised NLP Sequence Modeling

AI for Engineering

Applying intelligent learning systems to engineering challenges including health monitoring, IoT integration, and industrial automation using machine learning pipelines.

IoT + ML Health Monitoring Recommendation Industrial AI Smart Systems

Natural Language Processing

Building extractive-abstractive summarization systems, cross-lingual models for Bangla–English language pairs, and computational methods for social media event analysis.

Text Summarization Cross-Lingual NLP Social Media Analysis Event Forecasting Bangla NLP

Publications

Research Publications

Peer-reviewed journal articles and conference contributions.

01
A Machine Learning-Driven IoT-Enabled Activity Monitoring System for Chronic Disease Prediction and Personalized Diet Planning
Md Yasir Arafat et al.
ICETSD 2026 — International Conference on Emerging Technology in Science and Development, JUST, Bangladesh
2026 Conference → Poster IoT Machine Learning
02
An Efficient Convolution Neural Network-based Novel Framework for Potato Leaf Diseases Classification and Identification
Md Yasir Arafat, Sawrnali Khatun, Monishanker Halder, Mohammad Nowsin Amin Sheikh, Md Nasim Adnan, Sk Shalauddin Kabir
IAENG International Journal of Computer Science, Vol. 52(7), 2025
2025 Computer Vision Deep Learning
03
Automatic Text Summarization Based on Extractive-Abstractive Method
Md. Ahsan Habib, Romana Rahman Ema, Tajul Islam, Md. Yasir Arafat, Mahedi Hasan
Radioelectronic and Computer Systems, Vol. 106(2), 2023
2023 DOI NLP
04
Protein Secondary Structure Prediction Using Hybrid Recurrent Neural Networks
Romana Rahman Ema, A. Khatun, M. A. Hossain, Mostafijur Rahman Akhond, Nazmul Hossain, Md. Yasir Arafat
Journal of Computer Science, Vol. 18(7), pp. 599–611, 2022
2022 DOI Deep Learning Bioinformatics
05
Technical Challenges of Cloud Forensics and Suggested Solutions
Md. Yasir Arafat, Bipasha Mondal, Sreeti Rani
International Journal of Scientific and Engineering Research, Vol. 8(8), 2017
2017 DOI
06
JUSTCE: A Secured and Reliable E-Learning Portal
Md. Kamrul Islam, Md. Yasir Arafat, Ahsan-Ul Ambia
International Journal of Scientific and Engineering Research, Vol. 7(7), 2016
2016
07
Redundant Reduced LZW (RRLZW) Technique of Lossless Data Compression
Md. Kamrul Islam, Md. Yasir Arafat
International Journal of Advanced Research in Computer Science, Vol. 5(6), 2014
2014 Algorithms

Projects

Research Projects

Funded research initiatives and applied AI development work.

PI · Funded Research

IoT-Based Health Monitoring System with Fitness & Diet Recommendation

Designed and led a machine learning-driven health monitoring system integrating IoT sensors with personalized fitness and dietary recommendation algorithms. The system enables real-time health data collection and intelligent adaptive recommendations.

Funding Agency: Jashore University of Science and Technology
Grant Amount: 150,000 BDT (~$1,250 USD)
Period: July 2023 – June 2024
IoT Machine Learning Python Health AI Recommendation Systems
SPMT · HEQEP Funded

Integration of IT Facilities in Teaching-Learning Activities

Sub-Project Management Team member for a large-scale ICT integration initiative at the Department of Environmental Science and Technology, funded by the University Grant Commission of Bangladesh under the HEQEP project.

Funding Agency: HEQEP / University Grant Commission, Bangladesh
Grant Amount: 1.21 crore BDT (~$121,000 USD)
Period: May 2017 – April 2018
ICT e-Learning Educational Technology
Research · Computer Vision

Potato Leaf Disease Classification Framework

Developed a CNN-based framework for automated classification and identification of potato leaf diseases from digital images, contributing to AI-driven precision agriculture solutions.

Status: Published (IAENG IJCS, 2025)
CNN Computer Vision Python YOLOv8 Agricultural AI
Research · NLP

Cross-Language Text Summarization (Bangla–English)

Internship research project at IIT Kharagpur under Prof. Pabitra Mitra, focusing on cross-lingual extractive-abstractive summarization for Bangla–English language pairs, contributing to low-resource NLP research.

Supervisor: Prof. Pabitra Mitra, IIT Kharagpur
Period: December 2022 – January 2023
NLP Summarization Bangla Transformer Python
M.Sc. Thesis

Civil Unrest Event Forecasting via Social Media

Developed a novel ML pipeline to filter civil unrest-related Twitter content using keyword dictionaries and machine learning, then forecast future event dates, times, and locations using geolocation analysis on streaming data.

Supervisor: Engr. Md. Arif Rahman
Period: 2013 – 2015
Machine Learning NLP Twitter API Geolocation Event Detection
B.Sc. Thesis

Redundant Reduced LZW (RRLZW) Data Compression

Designed and evaluated a novel lossless data compression technique building on the LZW dictionary method, achieving improved compression ratios on textual data compared to standard approaches.

Supervisor: Md. Kamrul Islam, PhD
Period: 2009 – 2012
Algorithms Data Compression LZW Huffman

Experience

Professional Timeline

Academic and industry positions shaping a research career at the intersection of AI and education.

2024 – Present  Current
PhD Researcher (Doctorant)
CRIL – Centre de Recherche en Informatique de Lens, Université d'Artois, Lens, France
Doctoral thesis: "Apprentissage sur des objets 3D non-uniformes : une méthode basée sur les graphes" (Learning on Non-Uniform 3D Objects: A Graph-Based Method). Supervised by Prof. Saïd Jabbour and co-supervised by Dr. Wissem Inoubli. Research on GNNs for 3D mesh learning, scalable pooling strategies, and temporal GNN applications in fluid simulation and weather modelling. Funded by Université d'Artois. Research axis: Data Learning & Processing.
April 2018 – Present
Assistant Professor
Dept. of Computer Science & Engineering, Jashore University of Science and Technology
Teaching AI, Machine Learning, Pattern Recognition, Computer Vision, and Bioinformatics. Supervising undergraduate and Master's-level research projects and theses. Leading funded research on IoT and ML systems. Serving as BSc Examination Committee Chairman and ICT Cell Additional Director.
September 2014 – April 2018
Lecturer
Dept. of Computer Science & Engineering, Jashore University of Science and Technology
Taught Computer Fundamentals, Programming, Data Structures and Algorithms, Algorithm Analysis and Design, Digital Logic Design, and Microprocessor courses. Supervised undergraduate project and thesis work.
February 2014 – September 2014
.NET Programmer
SOFTWORKS Ltd.
Developed ERP systems using .NET technologies for the garment dyeing industry and built Medicine Warehouse Inventory Management Systems as part of an agile development team.
April 2013 – January 2014
Junior .NET Developer
AutoInformatik.AG
In-house software development and testing using ASP.NET and desktop application frameworks. Contributed to core business application development.

Recognition

Awards & Scholarships

Academic recognition, merit scholarships, and professional distinctions.

BDSET Training
2022 – 2023 · ICT Division, Bangladesh
Jashore Education Board Scholarship
2008 – 2012 · Jashore, Bangladesh
Grameen Bank Scholarship
April 2007 · For SSC Result, Bangladesh
Inter Disciplinary Debate Competition
2010 · Jashore University of Science and Technology

Dissemination

Talks & Conferences

Presentations, poster sessions, and conference participations in the research community.

2026

Conferences

Poster
A Machine Learning-Driven IoT-Enabled Activity Monitoring System for Chronic Disease Prediction and Personalized Diet Planning
ICETSD 2026 — International Conference on Emerging Technology in Science and Development, JUST, Bangladesh
2026
2025

Events

Talk
Scalable Graph Representation Learning
MAIA IA-Optim Workshop
4 June 2025
Talk
Thesis Hypothesis Presentation
JDD25 — Journée des Doctorants · Engaging with peers and experts in computer science
17 June 2025

Conferences

Poster
Learning 3D Molecular Representation: A Graph Transformer Approach
2SIC Conference — UPJV, Amiens
10–12 July 2025
Abstract
Scalable Graph Neural Network for Learning Mesh Graphs
NCCIA-2025 — National Conference on Computing and Information Automation, JUST, Jashore, Bangladesh
20 June 2025

Professional

Memberships

Active participation in professional engineering and research communities.

International Association of Engineers (IAENG), Hong Kong
Internet Society Bangladesh Dhaka Chapter
Société Informatique de France
JUST Teachers' Association, Jashore University of Science & Technology, Bangladesh
CSE Club, Dept. of Computer Science and Engineering, JUST, Bangladesh

Service

Peer Review Activity

Contributing to the scientific community as a reviewer for international journals and conferences.

7
Manuscripts Reviewed
3
Venues
2025
All in Year
Md Yasir Arafat serves as a peer reviewer for high-impact international journals and top-tier AI/ML conferences, including Springer's Journal of Supercomputing, IJCIA, and IJCNN 2025.
The Journal of Supercomputing
Springer · ISSN 0920-8542
3
Manuscripts Reviewed International peer-reviewed journal
2025
Recognised by certificate of contribution for reviewing 3 manuscripts in 2025.
International Journal of Computational Intelligence and Applications
IJCIA · World Scientific
1
Manuscript Reviewed International peer-reviewed journal
2025
Contributed 1 manuscript review in 2025 for this international computational intelligence journal.
International Joint Conference on Neural Networks 2025
IJCNN 2025 · IEEE · CMT
3
Articles Reviewed Top-tier IEEE neural networks conference
2025
CMT Portal →
Reviewed 3 conference articles for IJCNN 2025 via Microsoft CMT review system.

Skills

Technical Expertise

A broad toolkit spanning AI research, scientific computing, software development, and academic tools.

AI & Machine Learning

Deep Learning (CNNs, RNNs)
Graph Neural Networks
Computer Vision (YOLOv8, SVM)
NLP & Text Summarization
Recommendation Systems

Programming

Python
C# / .NET
C / C++
MATLAB
JavaScript / PHP / HTML5

Tools & Platforms

Jupyter Notebook
MATLAB / Simulink
Unity / Blender (XR)
Arduino / IoT
LaTeX

Databases & Web

MySQL
MS SQL Server
ASP.NET / MVC
Oracle / SQLite
IIS / XAMPP

Contact

Get In Touch

Open to research collaborations, academic discussions, and opportunities in GNN and AI research. Based in Lens, France.

arafatcse@just.edu.bd arafat.cse.just@gmail.com CRIL PhD Page — Université d'Artois JUST University Profile ResearchGate
Lens, Hauts-de-France, France