Saqib Nazir.

Working on the SCAnDi project (Single-cell and single molecule analysis for DNA identification), funded by the UKRI Economic and Social Research Council.

• Developing interpretable AI models for single-cell phenotyping in forensic and clinical applications
• Morphological classification of individual cells (sperm, skin, blood) and assessing cell integrity
• Integrating LLM-based interpretation to generate biologically grounded explanations
• Multi-modal analysis combining microscopy imaging with molecular data

Research Associate on the COSURIA project. Designed generative neural networks for geometry and color completion on 3D meshes of human scans for AR/VR/XR applications.

• Developed generative models for completing both geometry and color in 3D color meshes
• Consolidation of 3D scans of people to create 3D avatars for extended reality
• Addressed challenges in imperfect 3D data: gaps, acquisition errors, retouching

Assistant Researcher for the MENELAOS-NT European Training Network. Research focused on Deep Neural Networks for monocular depth estimation from defocus blur.

• Study of physical foundation for depth cues in images
• DNN-based solutions for depth inference from single-shot images exploiting defocus cues
• Definition of benchmarks for DNN training, validation and testing
• Evaluation of depth map accuracy using indoor and outdoor image collections

Research stay exploring LiDAR and TOF cameras to develop the iDFD dataset — an open-source dataset for Depth from Defocus for 3D indoor applications.

• Contribution to open-source projects for the team
• Development of the iDFD dataset for depth-from-defocus research
• Public presentation of the project and contributing to research papers

Collaboration with CiTIUS for the MENELAOS-NT project, developing 2HDED:NET for joint depth estimation and image deblurring from defocused images.

• Development of 2HDED:NET architecture
• Training and testing on NYU-Depth v2 and Make3D datasets
• Public presentation of results and contributing to research papers

Master's research at the Medical Imaging and Diagnostic Lab (MIDL) affiliated with the National Center of Artificial Intelligence (NCAI), developing CAD system solutions using Computer Vision and Deep Learning.

• Computer Aided Diagnostics (CAD) using Generative Adversarial Networks
• Master's thesis: "Generative Adversarial Networks for Enhancing Low-Dose CT Scans"
• Research in AI, Machine Learning, Deep Learning, and Computer Vision