Abdul Manaf

Email
abdulmanafsahito@gmail.com
Phone
+92 3003141736
Birthday
September 14, 2001
Location

Sukkur, Pakistan

About me

I'm Abdul Manaf — an AI researcher and engineer based in Gjøvik, Norway, specializing in computer vision, medical imaging, and Vision-Language Models (VLMs). I recently completed my MS Computer Science at Sukkur IBA University (CGPA: 3.45), where I ranked #1 in my MS batch and was awarded the PM's Youth Laptop Scheme 2025.

My MS thesis developed a ROI-guided multimodal framework for automated lesion description generation, benchmarking MedFlamingo, MedLLaVA, MedGemma, and Qwen2-VL across multiple fine-tuning strategies. During my degree, I completed a fully-funded 6-month research exchange at NTNU, Norway under the NORPART-CONNECT project, collaborating with the Intelligent Systems & Analytics (ISA) Research Group under Dr. Ali Shariq Imran — where I co-authored a publication on Swin-B Transformer chest X-ray classification and developed EduRAG, a RAG-based educational chatbot framework currently under review at Discover Artificial Intelligence.

I'm currently actively seeking PhD opportunities in AI, computer vision, or multimodal learning — open to positions across Europe and beyond. If you're a researcher or supervisor working in these areas, I'd love to connect. 🚀

What I'm doing

AI & Medical Imaging Research

Developing multimodal AI frameworks for medical image analysis — combining Vision-Language Models, ROI guidance, and RAG-based systems for healthcare applications.
LLM & RAG Systems Development

Building and benchmarking RAG pipelines using GPT, Gemini, LLaMA, Mistral & DeepSeek for domain-specific applications in education and healthcare.
Full-Stack AI Applications

Building end-to-end AI-powered web applications using React, Node.js, FastAPI, and cloud platforms — from research prototype to production deployment.
Enterprise & Government Software

Designing scalable backend systems and AI-integrated solutions for government-scale projects, including disability services platforms and complaint automation systems.

Key Achievements

🏆

#1 MS CS Batch

PM's Youth Laptop Scheme 2025

🇳🇴

NTNU Norway

Fully-funded NORPART-CONNECT Exchange

🎓

SEEF Scholarship

Govt. of Sindh — 4-year BS (2020)

📄

5+ Publications

IEEE Access, ScienceDirect, arXiv & more

🥈

RoboWar 2nd Place

University Techfest — Robotics & Engineering

GitHub Activity

🐙

GitHub Contributions

@AbdulManafSahito

View Full Profile on GitHub →

Resume

Education

SUKKUR IBA UNIVERSITY, Sukkur
MS(CS) Jan 2024 — Dec 2025
CGPA: 3.45 🏆 Ranked #1 in MS CS Batch PM's Youth Laptop Scheme 2025 (Hunarmand Pakistan) View Details

Key Coursework: Advanced NLP (A) · Advanced Machine Learning (A−) · Advanced Computer Vision · Advanced Cyber Security · Research Methodologies (A−) · Advanced Analysis of Algorithms

Master's Thesis:

"A Multimodal Framework to Generate the Lesion Description from ROI-Guided Medical Image Through Vision-Language Model"

Developed an ROI-guided multimodal pipeline comparing four Vision-Language Models — MedFlamingo, MedLLaVA, MedGemma, and Qwen2-VL — across full fine-tuning, LoRA, and zero-shot strategies. Evaluated on MedTrinity-25M using RAGAS metrics, uncovering architecture-dependent ROI effects and a critical semantic similarity–factual correctness gap in medical AI.

Supervisor: Dr. Ghulam Mujtaba Shaikh See profile
Co-Supervisor: Dr. Sher Muhammad Daudpota See profile

View Code

🌍 International Research Exchange:

Selected for a fully-funded 6-month Research Fellowship at the Norwegian University of Science and Technology (NTNU), Gjøvik, Norway — NORPART-CONNECT Project (Jan 2025 – Jul 2025), collaborating with the Intelligent Systems & Analytics (ISA) Research Group. View outcomes View exchange View Certificate
SUKKUR IBA UNIVERSITY, Sukkur
BS(CS) Aug 2019 — Aug 2023
CGPA: 3.08 Sindh Educational Endowment Fund (SEEF) Scholarship — BS 2020 (4-year) View Details

Key Coursework: Machine Learning · Artificial Intelligence · Computer Vision · Cyber Security · Parallel & Distributed Computing · Applied Robotics · Data Structures · Design & Analysis of Algorithms

Final Year Project / Thesis:

"Enhancing Wrist Abnormality Detection with YOLO: Analysis of State-of-the-Art Single-Stage Detection Models"

Built an automated pediatric wrist fracture detection system using YOLOv8x on the GRAZPEDWRI-DX dataset, addressing the scarcity of radiologists in diagnosing distal radius and ulna fractures in children and adolescents. Achieved fracture detection mAP of 0.95 and overall mAP of 0.77.

View Code Live Demo Paper Link

🥈 2nd Place — RoboWar Techfest · Competitive robotics & engineering challenge

Experience

Junior Developer — Sukkur IBA University (Sindh Government Projects)
Nov 2025 – Present · Remote, Gjøvik Norway
Working remotely from Gjøvik, Norway on government-scale digital solutions for Sindh province. Key contributions include: a complaint similarity module using AI techniques with AWS deployment; email automation workflows via n8n.io for sjp.org.pk candidate queries; and core development of the DEPD (Department of Empowerment of Persons with Disabilities) prototype — managing backend verification services, a notifications/announcements broadcasting module, and PostgreSQL database design.
Student Researcher — Norwegian University of Science and Technology (NTNU)
Jan 2025 – Jul 2025 · 7 mos · Gjøvik, Norway
Conducted research under the ISA (Intelligent Systems & Analytics) Research Group led by Dr. Ali Shariq Imran, as part of the fully-funded NORPART-CONNECT Project.

Research Output 1: Co-authored peer-reviewed publication — "Exploring the Impact of Self-Supervised Learning and Swin-B Transformer on Saliency Maps and Performance in Multi-Label Chest X-Ray Classification" — Swin-B SSL model achieved AUC 0.83 and IoU 0.16 on CXR pathology detection. Developed an open-source web app for clinician use. Paper Link

Research Output 2 (First Author): Developed EduRAG — a RAG-based educational chatbot framework benchmarking GPT-RAG, Gemini-RAG, LLaMA-RAG, Mistral-RAG & DeepSeek-RAG across three university courses. GPT-RAG achieved highest answer relevancy (0.88–0.98); Gemini-RAG led on faithfulness (0.72–0.84). Under Review — Discover Artificial Intelligence.
Project Page·Live Demo·Code

Utilized HPC environments at NTNU for large-scale LLM training and inference. Actively participated in ISA Research Group gatherings and collaborative activities, and explored Norwegian culture across Oslo, Lillehammer, Hamar, Gjøvik & Beitostølen.
Associate Researcher (Computer Vision Engineer) — CRAIB, Sukkur IBA University
Nov 2023 – May 2025 · 1 yr 7 mos · Sukkur, Pakistan
Contributed to the HEC-funded Post-Flood Disaster Management System project, developing deep learning and computer vision applications for UAV-based flood rescue operations and Visual Question Answering (VQA) for post-flood aerial image analysis. Built NLP models for hate speech detection with XAI (LIME) interpretation. Developed the Health IQ app integrating ChatGPT for medical symbol interpretation. Enhanced the Sensorkit robotics platform and mentored undergraduate students in FYPs and robotics initiatives.
Machine Learning Engineer — FasTech Systems
Jun 2023 – Nov 2023 · 6 mos · Sukkur, Pakistan
Contributed to multiple AI initiatives including an AI-powered book creation platform, an intelligent hotel menu recommendation system, and a real-time audio call sentiment analysis tool. Applied ML algorithms, data processing pipelines, and AI frameworks to deliver measurable improvements in user experience and operational efficiency.
Researcher — DeepNLP.ai
Nov 2022 – Sep 2023 · 11 mos · Hybrid
Member of a multinational Deep-NLP research group (Europe & Asia) led by university faculty and postdoc fellows. Collaborated on NLP research spanning sentiment analysis, text classification, machine translation, and chatbot development. Contributed to medical imaging research on automated wrist fracture detection. deep-nlp.net
Data Science Intern — CodeClause
Mar 2023 – May 2023 · 3 mos · Remote
Worked on wine quality prediction and stock market forecasting projects. Applied ML algorithms, exploratory data analysis, and preprocessing techniques to build predictive models, sharpening skills in data analysis, model building, and optimization.

My Skills & Technologies

🤖 Artificial Intelligence & Machine Learning

Deep Learning Computer Vision Medical Image Analysis Vision-Language Models (VLMs) Object Detection Machine Learning Natural Language Processing RAG Systems Large Language Models (LLMs) Neural Networks YOLO Models (v5–v10) CNNs Transformer Architectures Transfer Learning Self-Supervised Learning Model Fine-Tuning (LoRA / Full) Generative AI Reinforcement Learning

💻 Programming Languages

Python JavaScript SQL HTML/CSS C++ Java R MATLAB

🛠️ Frameworks & Libraries

PyTorch TensorFlow OpenCV Keras Scikit-learn Hugging Face Transformers LangChain Pandas / NumPy Matplotlib / Seaborn FastAPI Streamlit Gradio n8n.io Prisma ORM

☁️ Cloud & DevOps

AWS (S3, EC2, deployment) Google Cloud Platform Docker Git & GitHub Firebase HPC (NTNU) PostgreSQL MLflow

🌐 Web Development

React.js Node.js Express.js Bootstrap MongoDB MySQL REST APIs Responsive Design

📚 Research & Academic

Research Methodology Academic Writing Systematic Literature Review (PRISMA) Experimental Design Statistical Analysis RAGAS Evaluation Conference Presentations Peer Review

🎯 Specialized Domains

Medical Imaging Multimodal AI Fracture Detection Pediatric Healthcare AI Educational AI & Chatbots Disaster Management Systems UAV / Drone Technology Explainable AI (XAI) Saliency Maps & GradCAM

🧠 Core Competencies

Problem Solving Critical Thinking Algorithm Design Team Collaboration Mentoring Project Management Technical Documentation Agile Methodology

CV

Publications

Journal Articles

Fake Reviews Detection on E-Commerce Websites Using Novel User Behavioral Features: An Experimental Study
2025
Authors: N Mughal, G Mujtaba, MH Mughal, A Manaf, Z Kamangar. Published in ACM Transactions on Asian and Low-Resource Language Information Processing.

Abstract: The trend of writing fake reviews has recently increased with the rapid growth of e-commerce websites. Fake reviews are usually written to promote or demote the targeted products to affect the customer's decision and thus achieve a competitive advantage. Several techniques have been proposed to detect fake reviews written in English, and promising results have been obtained in the literature. Nevertheless, detecting fake reviews for low-resource languages (such as Roman Urdu) is still in the infancy stage and suffers from low classification results for two main reasons. Firstly, the existing studies mostly worked on textual features or lingual features. Secondly, the datasets used in existing studies are highly imbalanced, and proper attention to this issue may further enhance the performance. Therefore, to address these weaknesses and further enhance the performance, we have identified three types of discriminative features: review textual features, review lingual features, and review behavioral features using the Daraz dataset. Moreover, we evaluated LSTM-based text generation techniques for textual features and the random undersampling and oversampling for behavioral and lingual features to deal with class imbalance problems. Finally, we empirically evaluated the performance of machine learning and deep learning algorithms in classifying fake reviews written in the Roman Urdu language. The experimental results show that user behavioral features play a vital role in detecting fake reviews. Moreover, it was found that text generation is ineffective for balancing the textual data because the informative feature for fake review detection depends on the user behavioral features compared to textual features. Finally, the experimental results show that gradient boosting (GB) outperformed other models and improved 3% accuracy from the baseline study.

View Code
Aerial Image Classification in Post Flood Scenarios Using Robust Deep Learning and Explainable Artificial Intelligence
2025
Authors: A Manaf, N Mughal, K R Talpur, B A Talpur, G Mujtaba, S R Talpur. Published in IEEE Access.

Abstract: Efficiently delivering timely assistance to flooded regions is a critical imperative, and leveraging deep-learning methodologies has demonstrated significant efficacy in addressing environmental challenges. Several authors have collected data from specific regions and presented the methodologies for classifying flooded images. However, there are two main limitations of existing methodologies and benchmark datasets. Firstly, the models are trained on the images collected from specific geographical regions, which limits their ability to generalize when encountering images with diverse features or from varied regions. Secondly, the models are trained on high-resolution images and lack classification for the low-resolution images. In this study, we curated a dataset by merging benchmark datasets and acquiring images from web repositories. The main objective of this study is to address resolution issues and enhance model accuracy across diverse regions. We conducted a comparative analysis of the curated dataset using various deep-learning models based on CNN architecture. The experimental findings revealed that MobileNet and Xception outperformed ResNet-50, VGG-16, and Inception(v3) models, achieving an impressive accuracy rate of approximately 98% and an f1-score of 92% for flood class. Additionally, we have also used XAI Lime to interpret model results.

Project Page View Code
Exploring the Impact of Self-Supervised Learning and Swin-B Transformer on Saliency Maps and Performance in Multi-Label Chest X-Ray Classification
2025
Authors: A Manaf (co-author), AS Imran, G Mujtaba, et al. Published in IEEE. Research conducted at NTNU (NORPART-CONNECT) with the ISA Research Group.

Abstract: This study explores self-supervised learning (SSL) and Swin-B transformer architectures for multi-label chest X-ray (CXR) classification, with a focus on saliency maps and model interpretability. We compared Vision-Language Models and ROI-guided pipelines on medical imaging. The Swin-B SSL model achieved AUC 0.83 and IoU 0.16 on CXR pathology detection. An open-source web application was developed for clinician use.

Paper View Code
Enhancing Wrist Abnormality Detection with YOLO: Analysis of State-of-the-Art Single-Stage Detection Models
2024
Authors: A Ahmed, AS Imran, A Manaf, Z Kastrati, SM Daudpota. Published in Biomedical Signal Processing and Control 93, 106144.

Abstract: Diagnosing and treating abnormalities in the wrist, specifically distal radius, and ulna fractures, is a crucial concern among children, adolescents, and young adults, with a higher incidence rate during puberty. However, the scarcity of radiologists and the lack of specialized training among medical professionals pose a significant risk to patient care. This problem is further exacerbated by the rising number of imaging studies and limited access to specialist reporting in certain regions. This highlights the need for innovative solutions to improve the diagnosis and treatment of wrist abnormalities. Automated wrist fracture detection using object detection has shown potential, but current studies mainly use two-stage detection methods with limited evidence for single-stage effectiveness. This study employs state-of-the-art single-stage deep neural network-based detection models YOLOv5, YOLOv6, YOLOv7, and YOLOv8 to …

Project Page View Code

Conference Articles

Small Data, Big Impact: A Multi-Locale Bone Fracture Detection on an Extremely Limited Dataset Via Crack-Informed YOLOv9 Variants
2024
Authors: A Ahmed, AS Imran, A Manaf, Z Kastrati. Presented at the 21st International Conference on Frontiers of Information Technology (FIT 2024), Islamabad, Pakistan, December 9-10.

Abstract: Automated wrist fracture recognition has become a crucial research area due to the challenge of accurate X-ray interpretation in clinical settings. This study utilizes an extremely limited multi-region fracture dataset and proposes a novel approach where YOLOv9 is pre-trained on surface cracks rather than COCO. This method achieves state-of-the-art (SOTA) performance on the newly released FracAtlas dataset, improving the mean average precision (mAP) score by 7% and sensitivity by 13%.

Project Page View Code

ArXiv/Preprint Papers

AI-Enhanced Pediatric Pneumonia Detection: A CNN-Based Approach Using Data Augmentation and Generative Adversarial Networks (GANs)
2025
Authors: A Manaf, N Mughal. Preprint available at arXiv:2507.09759.

Abstract: Pneumonia is a leading cause of mortality in children under five, requiring accurate chest X-ray diagnosis. This study presents a machine learning-based Pediatric Chest Pneumonia Classification System to assist healthcare professionals in diagnosing pneumonia from chest X-ray images. The CNN-based model was trained on 5,863 labeled chest X-ray images from children aged 0-5 years from the Guangzhou Women and Children's Medical Center. To address limited data, we applied augmentation techniques (rotation, zooming, shear, horizontal flipping) and employed GANs to generate synthetic images, addressing class imbalance. The system achieved optimal performance using combined original, augmented, and GAN-generated data, evaluated through accuracy and F1 score metrics. The final model was deployed via a Flask web application, enabling real-time classification with probability estimates. Results demonstrate the potential of deep learning and GANs in improving diagnostic accuracy and efficiency for pediatric pneumonia classification, particularly valuable in resource-limited clinical settings.

View Code
Pediatric Wrist Fracture Detection in X-rays via YOLOv10 Algorithm and Dual Label Assignment System
2024
Authors: A Ahmed, A Manaf. Preprint available at arXiv:2407.15689.

Abstract: Wrist fractures are highly prevalent among children and can significantly impact their daily activities, such as attending school, participating in sports, and performing basic self-care tasks. If not treated properly, these fractures can result in chronic pain, reduced wrist functionality, and other long-term complications. Recently, advancements in object detection have shown promise in enhancing fracture detection, with systems achieving accuracy comparable to, or even surpassing, that of human radiologists. The YOLO series, in particular, has demonstrated notable success in this domain. This study is the first to provide a thorough evaluation of various YOLOv10 variants to assess their performance in detecting pediatric wrist fractures using the GRAZPEDWRI-DX dataset. It investigates how changes in model complexity, scaling the architecture, and implementing a dual-label assignment strategy can enhance detection performance. Experimental results indicate that our trained model achieved mean average precision (mAP@50-95) of 51.9%, surpassing the current YOLOv9 benchmark of 43.3% on this dataset. This represents an improvement of 8.6%.

Project Page View Code

Projects

🚀 Live Demo 📁 GitHub

Student AI Navigator - Intelligent RAG-based Educational Chatbot

Agentic AI
🚀 Live Demo 📁 GitHub

LinkedInCraft - AI-Powered LinkedIn Post Generator

Agentic AI
🚀 Live Demo 📁 GitHub

AI Finance Agent Team - Multi-Agent Financial Intelligence

Agentic AI
🚀 Live Demo 📁 GitHub

AI Research Agent - Professional Research Reports Powered by GPT-4

Agentic AI
🚀 Live Demo 📁 GitHub

PromptFlow Studio

Agentic AI
🚀 Live Demo 📁 GitHub

Vision OCR

Computer Vision
🚀 Live Demo 📁 GitHub

Pediatric Chest Pneumonia Classification

Computer Vision
🚀 Live Demo 📁 GitHub

Pediatric Wrist Abnormality Detection System

Computer Vision
Twitter Sentiment Analysis

Computer Vision
🚀 Live Demo 📁 GitHub

CortexTest: LLM Assessment Tool

Machine Learning and AI
From Text to Diagnosis

Machine Learning and AI
YOLOv10 Fracture Detection

Machine Learning and AI
Book Generation with GPT-3.5 Turbo

Machine Learning and AI
Finetuning with LoRA for Stable Diffusion

Machine Learning and AI
🚀 Live Demo 📁 GitHub

Premier Bank

Web Development
🚀 Live Demo 📁 GitHub

QR Code Generator

Web Development
Recipe Management System

Web Development
BookStore MERN Project

Web Development
🚀 Live Demo 📁 GitHub

SensorKit Website

Web Development
NMR Text Editor

Software Development
Mini Compiler with GUI

Software Development
DQA Chatbot

NLP and Chatbots
Tweets Sentiment Analysis

NLP and Chatbots

Achievements

Norwegian University of Science and Technology (NTNU), Gjøvik Campus

Beautiful view of Oslo, Norway during the exchange program

With fellow researchers and team members at NTNU

Cultural exchange dinner with international researchers

Completed My 6-Month Fully Funded Research Exchange at NTNU, Norway! ✅🌍

Jan 2025 – Jul 2025 📍 Gjøvik, Norway

I was selected as a Research Fellow at the Norwegian University of Science and Technology (NTNU) in Gjøvik, Norway, as part of the NORPART-CONNECT Project! 🙌

During the six-month exchange, I dove deep into cutting-edge research in Artificial Intelligence (AI), Machine Learning (ML), Deep Learning (DL), and Large Language Models (LLMs). My focus spanned multiple fields, including education, medical imaging, and environmental applications, harnessing AI to solve real-world challenges. 📚💡

🤝 Collaborating with Leading Experts

The opportunity allowed me to collaborate with top-tier researchers, immerse myself in cutting-edge AI methodologies, and contribute to impactful projects that connect technology and education — including developing AI-powered educational tools and fostering global research collaborations. 🌐✨

Thank you to Sukkur IBA University and the Norwegian University of Science and Technology (NTNU) for this incredible opportunity. I'm deeply grateful for the unwavering support, guidance, and mentorship from my professors, supervisors, and supporters: Sher Muhammad Daudpota, Ali Shariq Imran, and Ghulam Mujtaba Shaikh, whose trust and guidance have been instrumental in getting me to this point. 🙏💙

Grateful for the journey of learning, networking, and contributing to the future of AI in education! 🎓🤖

#AI #MachineLearning #DeepLearning #NLP #LLMs #EdTech #Research #Education #NTNU #Norway #NORPART #AIforEducation

Laptop received under PM's Youth Laptop Scheme 2025 (Hunarmand Pakistan)

Distribution ceremony with fellow recipients

PM's Youth Laptop Scheme 2025 — Hunarmand Pakistan

Received Laptop under PM's Youth Laptop Scheme 2025 (Hunarmand Pakistan) 💻🇵🇰

2025 📍 Pakistan

I'm grateful to have been selected as a recipient of the PM's Youth Laptop Scheme 2025, under the Hunarmand Pakistan initiative. This government scheme aims to equip talented students and youth with the tools needed to excel in education, research, and digital skills. 🙌 Official scheme website →

The laptop will support my ongoing work in AI research, machine learning, and academic projects during my MS program at Sukkur IBA University. Thank you to the Government of Pakistan and all those involved in making this initiative possible for students across the country. 📚✨

🎓 Empowering Youth Through Technology

Initiatives like the PM's Youth Laptop Scheme help bridge the digital divide and enable students to participate in research, online learning, and innovation. I'm committed to using this resource to contribute to my field and give back to the community.

Grateful for this opportunity — here's to learning, building, and growing with the support of Hunarmand Pakistan! 💙

#PMLaptopScheme #HunarmandPakistan #YouthEmpowerment #Pakistan #Education #DigitalPakistan #StudentLife #SukkurIBA

Government of Sindh — Sindh Educational Endowment Fund (SEEF) Scholarship Program

Sindh Educational Endowment Fund (SEEF) Scholarship — Four-Year BS Support 🎓

2020 – 2024 📍 Sindh, Pakistan

I was awarded the Sindh Educational Endowment Fund (SEEF) scholarship in BS year 2020, under the Government of Sindh, College Education Department. The scholarship supported my entire four-year undergraduate (BS) studies. 🙌

📜 About SEEF

Sindh Educational Endowment Fund is registered vide Trust Deed No. 625/2020. The program provides need-based scholarships to meritorious students in Sindh for higher education. I am grateful to the Government of Sindh and the College Education Department for this opportunity, which enabled me to pursue my BS in Computer Science at Sukkur IBA University.

Official SEEF portal (seef.sindh.gov.pk) → · SEEF Scholarship (form-seef.com) →

Thank you to the Government of Sindh and SEEF for supporting my education — here's to continued learning and giving back! 💙

#SEEF #SindhGovernment #Scholarship #Education #Sindh #Pakistan #StudentLife #SukkurIBA