Imran Razzak
Associate Professor, MBZUAI (Department of Computational Biology) • Personalized Medicine • Medical Imaging • Multimodal Clinical Reasoning
Hello. Hola. Aloha. Bonjour. स्वागत , مرحباً, ښه راغلاست , خوش آمدید, Haai, Grüßgott. Goedendag. Oi. Shalom. Assalam o Alaikum, Guten Tag. Ciào. Malo e lelei. Hei. 欢迎, 안녕. Yo.
Imran Razzak
Associate Professor, Computational Biology
Datasets & benchmarks
We build and release datasets that enable trustworthy and generalizable AI, spanning medical imaging, multimodal learning, and real-world deployment settings. Below are a few representative datasets and demos from our past work.
Publications
For the most up-to-date publication list, please visit Google Scholar .
Our Datasets
A curated selection of datasets and benchmark resources (images shown are illustrative previews).
Marine-tree: A Large-scale Hierarchically Annotated Dataset for Marine Organism Classification
Hierarchical image classification • fine-grained taxonomy • large-scale labeling
Jarvis: A Voice-based Context-as-a-Service Mobile Tool for a Smart Home Environment
Context-as-a-service • mobile sensing • voice interaction
Brain Shift Correction: Intra-operative Ultrasound (iUS)
Surgical guidance • segmentation • registration / correction
Cardiac MRI under Respiratory Motion
Motion robustness • cardiac analysis • temporal consistency
Social Media Event Detection (Demo/Benchmark)
Event detection • multimodal / social signals • real-world streams
Hierarchical Classification (Additional Benchmark)
Taxonomy-aware learning • long-tailed categories • robust evaluation
