Anastasios Yiannakidis

Max Planck Institute for Intelligent Systems · Tübingen, Germany

Research Software Engineer at the Max Planck Institute for Intelligent Systems, building scalable pipelines for synthetic data generation, multi-view markerless motion capture, and 3D human and animal reconstruction.

Motion Capture 3D Animation Synthetic Data Graphics Deep Learning Computer Vision

Download CV

Experience

Research Engineer

May 2022 - Present

Max Planck Institute for Intelligent Systems

Perceiving Systems Department

Tübingen, Germany

Research Engineer

Sep 2019 - May 2022

University of Cyprus

Graphics & Extended Reality Lab

Nicosia, Cyprus

BSc Computer Science

Sep 2015 - Jun 2019

University of Cyprus

8.65 / 10 - Student Excellence Award. 30 ECTS Erasmus at Linnaeus University, Sweden.

Nicosia, Cyprus

Research Engineer (Intern)

Jun 2018 - Sep 2018

CYENS - Centre of Excellence

Monocular 3D human motion reconstruction (thesis project)

Nicosia, Cyprus

Technical Skills

Programming

Python C++ C# C Java JavaScript PHP Bash SQL Linux Docker Apptainer Conda HTCondor Git

ML & Computer Vision

PyTorch PyTorch3D NumPy Pandas OpenCV scikit-learn Synthetic Data Optimization Calibration Weights & Biases

3D & Graphics

Unreal Engine Blueprints Blender Unity Rendering Animation Rigging Mesh Processing Point Clouds Depth & Normal Maps

Motion Capture

Markerless Multi-View Retargeting Vicon Shogun PhaseSpace Optitrack Autodesk MotionBuilder

Human Modeling

SMPL-X FLAME MANO 3D Body Pose & Shape Estimation Facial Expression Eye-Gaze

Web

React Node.js Laravel Drupal HTML CSS Bootstrap jQuery REST APIs MySQL phpMyAdmin

Publications

Oral Presentation

MAMMA: Markerless Accurate Multi-person Motion Acquisition

H. Cuevas-Velasquez, A. Yiannakidis, S. Shin, G. Becherini, M. Höschle, J. Tesch, T. Obersat, T. Alexiadis, E. Halilaj, M.J. Black

We introduce MAMMA, a novel markerless motion-capture pipeline that reconstructs accurate SMPL-X pose and shape directly from multi-view video. By predicting dense 2D surface landmarks and handling complex multi-person interactions and occlusions, MAMMA achieves reconstruction quality competitive with commercial marker-based systems.

CVPR 2026

Paper Project

Best Paper Runner-Up Oral Presentation

NeuralFur: Animal Fur Reconstruction From Multi-View Images

V. Sklyarova, B. Kabadayi, A. Yiannakidis, G. Becherini, M.J. Black, J. Thies

Reconstructing animal fur from images is difficult due to its fine detail and lack of training data. This work introduces a multi-view approach that uses VQA to infer fur properties and guide strand growth on top of coarse geometry, resulting in realistic 3D fur across different animals. The representation is compatible with existing game and physics engines, enabling artists to simulate and manipulate fur within standard production pipelines.

3DV 2026

Paper Project Code

Oral Presentation

BEDLAM2.0: Synthetic Humans and Cameras in Motion

J. Tesch, G. Becherini, P. Achar, A. Yiannakidis, M. Kocabas, P. Patel, M.J. Black

BEDLAM 2.0 is a large-scale synthetic video dataset of animated humans in simulated clothing, built for 3D pose and shape estimation. With over 8 million images, it expands the original BEDLAM with greater pose and shape diversity, added shoes and strand-based hair, and a broad range of realistic cameras and motions.

NeurIPS 2025

Paper Project Code

Generative Zoo

T. Niewiadomski, A. Yiannakidis, H. Cuevas-Velasquez, S. Sanyal, M.J. Black, S. Zuffi, P. Kulits

We propose a pipeline for the scalable generation of realistic 3D animal pose and shape estimation training data. Training solely on data produced using our pipeline, we achieve state-of-the-art performance on a real-world 3D pose and shape estimation benchmark.

ICCV 2025

Paper Project Code

Rhythm is a Dancer: Music-Driven Motion Synthesis with Global Structure

A. Aristidou, A. Yiannakidis, K. Aberman, D. Cohen-Or, A. Shamir, Y. Chrysanthou

We present a music-driven neural framework that generates realistic human motions, which are rich, avoid repetitions, and jointly form a global structure that respects the culture of a specific dance genre. We illustrate examples of various dance genres, demonstrating choreography control and editing in a number of applications.

IEEE TVCG 2022 · 29(8)

Paper ArXiv Video Lab

Digitizing Wildlife: The Case of a Reptile 3-D Virtual Museum

S. Zotos, M. Lemonari, M. Konstantinou, A. Yiannakidis, G. Pappas, P. Kyriakou

We design and develop a 3D virtual museum with holistic metadata documentation and a variety of captured reptile behaviors and movements. Our main contribution lies on the procedure of rigging, capturing, and animating reptiles, as well as the development of a number of novel educational applications.

IEEE CGA 2022 · 42(5)

Paper Project

Virtual Dance Museum: The Case of Greek/Cypriot Folk Dancing

A. Aristidou, N. Andreou, L. Charalambous, A. Yiannakidis, Y. Chrysanthou

This paper presents a virtual dance museum developed to allow for widely educating the public, most specifically the youngest generations, about the story, costumes, music, and history of our dances. The museum is publicly accessible, and also enables motion data reusability, facilitating dance learning applications through gamification.

GCH 2021

Paper Project Database

Real-Time 3D Human Pose and Motion Reconstruction from Monocular RGB Videos

A. Yiannakides, A. Aristidou, Y. Chrysanthou

We fit 2D estimated poses, extracted from a single camera via OpenPose, with a 2D multiview joint projections database that is associated with their 3D motion representations. We then retrieve the 3D body pose of the tracked character, ensuring throughout that the reconstructed movements are within a feasible set.

CASA 2019 · 30(3-4)

Paper Video Code