Hamza Shafiq

I am a Doctoral Researcher at TU/e, supervised by Dr. Aaqib Saeed. My research focuses on Efficient Multi-Modal Learning, specifically addressing the challenges of efficient AI in clinical and non-clinical settings.

Previous Experience

Previously, I was a Senior Research Engineer (AI) in South Korea, where I focused on:

• Generative AI: Image restoration, colorization, and enhancement using adversarial networks and transformers.
• Computer Vision: Detection, tracking, segmentation, and restoration for security and surveillance applications.
• Hardware-Aware Optimization: Model optimization using TensorRT and CUDA for edge deployment.

During my master’s, I worked on restoration and colorization of old images using adversarial networks and dental image segmentation for automated diagnostics.

Technical Skills & Interests

• Deep Learning & Efficient AI: State Space Models (Mamba), Transformers, GANs, Stable Diffusion, Neural ODEs
• Multi-Modal Learning: Video understanding, physiological signal processing, clinical AI
• Computer Vision: Tracking, detection, segmentation, restoration, super-resolution
• Software & System Development: PyTorch, TensorFlow, JAX, TensorRT, CUDA, FastAPI, Gradio, Docker