Publications
Journal articles, conference papers, and preprints on medical imaging, computer vision, and language models. Q1/Q2 refer to journal quartile ranking.
Under review
-
First authorUnder review
EduRAG: An LLM-Powered Retrieval-Augmented Framework for Educational Applications
Submitted to Discover Artificial Intelligence (Springer Nature)
Benchmarks GPT, Gemini, LLaMA, Mistral, and DeepSeek RAG pipelines across three university courses. GPT-RAG led on answer relevancy (0.88–0.98); Gemini-RAG on faithfulness (0.72–0.84).
2026
-
JournalQ1Do Language Families Matter? Evaluating LLMs for Sentiment Analysis Through a Hierarchical Cross-Lingual Lens
Frontiers in Artificial Intelligence, vol. 9, 1854873
-
JournalQ1Exploring the Impact of Self-Supervised Learning and Swin-B Transformer on Saliency Maps and Performance in Multi-Label Chest X-Ray Classification
IEEE Access
Work with the ISA Research Group at NTNU. Swin-B with self-supervised pre-training reached AUC 0.83 on multi-label pathology detection.
2025
-
First authorJournalQ1
Aerial Image Classification in Post-Flood Scenarios Using Robust Deep Learning and Explainable Artificial Intelligence
IEEE Access, vol. 13, pp. 35973–35984
MobileNet and Xception reached about 98% accuracy and a 0.92 F1-score on the flood class; LIME explanations for model decisions.
-
JournalQ2Fake Reviews Detection on E-Commerce Websites Using Novel User Behavioral Features: An Experimental Study
ACM Transactions on Asian and Low-Resource Language Information Processing
-
First authorPreprintAI-Enhanced Pediatric Pneumonia Detection: A CNN-Based Approach Using Data Augmentation and Generative Adversarial Networks (GANs)
arXiv:2507.09759
2024
-
JournalQ1Enhancing Wrist Abnormality Detection with YOLO: Analysis of State-of-the-Art Single-Stage Detection Models
Biomedical Signal Processing and Control (Elsevier), vol. 93, 106144
YOLOv8 on GRAZPEDWRI-DX: 0.95 mAP on fractures, 0.77 mAP overall.
-
Conference
Small Data, Big Impact: A Multi-Locale Bone Fracture Detection on an Extremely Limited Dataset via Crack-Informed YOLOv9 Variants
International Conference on Frontiers of Information Technology (FIT), pp. 1–6
-
No publications match your search.
Theses
Degree research
-
MS Thesis · 2025A Multimodal Framework to Generate the Lesion Description from ROI-Guided Medical Image Through Vision-Language Model
Sukkur IBA University · Supervisors: Dr. Ghulam Mujtaba Shaikh, Dr. Sher Muhammad Daudpota
ROI-guided pipeline comparing MedFlamingo, MedLLaVA, MedGemma, and Qwen2-VL under full fine-tuning, LoRA, and zero-shot settings on MedTrinity-25M, evaluated with RAGAS.
-
BS Final Project · 2023
Enhancing Wrist Abnormality Detection with YOLO: Analysis of State-of-the-Art Single-Stage Detection Models
Sukkur IBA University · Later published in Biomedical Signal Processing and Control