Research Papers | Arabic NLP, OCR & Medical AI Publications
Academic publications on Arabic OCR, medical AI, federated learning, and efficient NLP — bridging research and production for under-represented languages.
My research focuses on making AI work for languages and domains that don’t get enough attention: Arabic text recognition, explainable medical imaging, and efficient models that run on limited hardware. I’ve published at IEEE Access, Scientific Reports (Nature), NeurIPS, and Springer, with work spanning from federated fine-tuning of LLMs to infant cry analysis.
Below is a summary of each paper with links to the full text.
Arabic Language Technology
QARI-OCR: High-Fidelity Arabic Text Recognition through Multimodal LLM Adaptation
arXiv
Healthcare and Medical AI
Explainable AI for Alzheimer’s Detection via Particle Swarm Optimization
ResearchGate
Explainable Self-Supervised Learning for Medical Image Diagnosis Based on DINO V2
Nature
Optimizing Medical Image Classification with Advanced Segmentation
ResearchGate
AI-Powered ICare: Infant Cry Analysis and Speech Therapy
ResearchGate
Leverage DINOv2 on Parasitic Egg Detection
Springer
Federated Learning & LLMs
Flowertune: A Cross-Domain Benchmark for Federated Fine-Tuning of LLMs
arXiv