Profile banner
Nikolaos Zioulis

Nikolaos Zioulis

CTO at Moverse — computer vision, graphics & machine learning

nikoz
Thessaloniki, Greece
Joined September 2025
Favicon imagezokin.github.io
Favicon imagegithub.com
Favicon imageorcid.org
Favicon imageportalcientifico.upm.es

Network

4.1K connections
🌐
XR & Immersive Experience Professionals
💻
AI & Software Engineers
🎨
Creative Technologists & Virtual Production
🚀
Tech Startup Founders & Leaders
🎓
Academic & Research Professionals
🎬
Technical & Creative Directors
👨‍💻
...

Summary

Bridges academic research and product engineering: Nikolaos has a sustained academic publication record (monocular depth, volumetric capture, motion capture and related methods) and translates that research into commercial products and company R&D as CTO of Moverse. github+2
Established leader in volumetric capture and markerless motion capture: across a decade at CERTH and in startup roles he has led technical work-packages (H2020 projects) and designed low‑cost volumetric capture systems—now focusing on scalable, real‑time MoCap products at Moverse. github+2
Practical open-source and reproducible research mindset: maintains an active GitHub presence (nmvrs), publishes project pages and code for multiple papers (e.g., KBody), and emphasizes datasets, benchmarks and project tooling in his publications. github+2
Productization and fundraising experience as a founding CTO: involved in Moverse’s product strategy and investor communications (e.g., TECS Capital announcement) and in collaborative EU projects (SPIRIT), indicating experience balancing research, product development and business needs. tecs+2

Work

Education

Projects

Writing

KBody: Towards general, robust, and aligned monocular whole-body estimation

April 1, 2023

A predict-and-optimize approach for monocular whole-body fitting that leverages virtual joints, disentangled optimization for pose/shape, asymmetric distance fields and generative model inversion as an appearance prior.

Favicon imagearxiv.org

Hybrid Skip: A Biologically Inspired Skip Connection for the UNet Architecture

May 1, 2022

Proposal of a biologically inspired skip connection variant for UNet architectures to improve learning dynamics and performance in imaging tasks (IEEE Access).

Favicon imagezokin.github.io

Monocular spherical depth estimation with explicitly connected weak layout cues

January 1, 2022

Work on monocular spherical panorama depth estimation that integrates weak layout cues to improve depth prediction for spherical imagery (ISPRS Journal of Photogrammetry & Remote Sensing).

Favicon imagesciencedirect.com