Yannis Douros

Computer Vision & Machine Learning Engineer
Applied R&D · Algorithms · 3D Geometry · C++ · Python

Software engineer with a background in computer vision, 3D geometry, image processing, machine learning and numerical computing.

My work has ranged from 3D human-body modelling and real-time facial tracking to indoor 3D reconstruction, deep learning and GPU-accelerated numerical computation. I am particularly interested in difficult back-end engineering problems: developing algorithms for problems that are not yet fully solved, and improving existing solutions against demanding performance targets.

Current interests and technologies

Computer Vision · Machine Learning · 3D Reconstruction · Image Processing
C++ · Python · OpenCV · PyTorch · OpenCL
Parallel/GPU Computing · Numerical Methods · SfM · NeRF

Currently open to new opportunities, including permanent and contract roles in computer vision, machine learning and applied R&D.

Current Projects

FractalKitchen
GPU-accelerated fractal rendering in modern C++ and OpenCL, exploring numerical precision, parallel computation and rendering techniques.
View on GitHub →

2D Gaussian Splatting
Experimental Python/PyTorch implementation exploring Gaussian splatting, differentiable rendering and image reconstruction in a simplified 2D setting.
View on GitHub →

---