CV
Curriculum vitae of Athanasios Papaioannou.
Contact Information
| Name | Athanasios Papaioannou |
| apapaion at gmail dot com |
Experience
-
2026 - Present London, UK
Research Software Engineer, Generative AI
Google
- Research and engineering on generative AI for 3D/4D vision applications.
-
2020 - 2026 London, UK
Senior Research Engineer
Huawei Research
- Led computer vision research on neural rendering, 3D/4D reconstruction, and generative modelling of humans and scenes, including feed-forward Gaussian splatting for novel-view synthesis.
- Designed and trained large-scale generative 3D models (statistical, mesh-based, and implicit morphable representations) for high-fidelity human modelling.
- Built end-to-end data pipelines for multi-view, RGB-D, and 4D capture data, supporting model training, evaluation, and deployment on complex visual datasets.
- Drove evaluation and ablation methodology for reconstruction quality, geometric accuracy, and view-consistency, enabling systematic model comparison and iteration.
- Translated research prototypes into integration-ready systems in collaboration with infrastructure and engineering teams; supervised junior researchers and contributed to research direction setting.
- Co-organised the View Synthesis Challenge for Human Heads at ICCVW 2023.
-
2019 - 2020 London, UK
Senior Research Engineer
Facesoft
- Developed deep learning models for face analysis, recognition, and representation learning, deployed in real-time production computer vision systems.
- Designed data processing and training pipelines for large-scale face datasets, with emphasis on robustness across demographics and capture conditions.
-
2018 - 2019 London, UK
Research Associate
Institute of Child Health, UCL
- Developed statistical and deep learning methods for analysing complex 3D craniofacial datasets, including mesh autoencoders for clinical 3D shape analysis.
- Built data processing and evaluation workflows for clinical 3D datasets and collaborated with surgeons, clinicians, and researchers on translational ML projects.
- Contributed to peer-reviewed publications on 3D shape modelling and clinical decision support.
-
2017 - 2018 London, UK
Research Associate
Imperial College London
- Developed face tracking and verification systems for secure online assessment platforms, including identity verification and behavioural analysis components deployed at scale.
-
2012 - 2016 London, UK
Research Assistant
Imperial College London
- Built large-scale 4D facial datasets for emotion and behaviour recognition research; developed algorithms for automated facial analysis and statistical 3D modelling.
- Contributed to national and European research projects on face and behaviour analysis.
Education
-
2012 - 2016 London, UK
PhD
Imperial College London
Computing
-
2003 - 2005 Thessaloniki, Greece
MSc
Aristotle University of Thessaloniki
Informatics
-
1999 - 2003 Thessaloniki, Greece
BSc
Aristotle University of Thessaloniki
Informatics
Interests
World Models:
Vision-Language-Action (VLA) Systems:
Agentic Perception:
Dynamic Scene Understanding:
Embodied AI:
Skills
3D / 4D Vision: Neural rendering, Gaussian splatting (3DGS, feed-forward variants), neural fields, differentiable rendering, novel-view synthesis, 3D/4D reconstruction, statistical and morphable shape models, mesh-based representations
Generative & Representation Learning: Autoencoders, self-supervised learning, multimodal representation learning, foundation models for vision
ML Systems & Methodology: Large-scale dataset construction, distributed and multi-GPU training, evaluation and ablation design, experimentation frameworks, research-to-production workflows
Programming: Python, C++ (production); previously MATLAB
Frameworks & Tools: PyTorch (primary), NumPy, scikit-learn, Git, Linux, Slurm, GPU/cluster computing
Publications
-
2026 ICo3D: An Interactive Conversational 3D Virtual Human
*joint first authorship
-
2025 imHead: A Large-Scale Implicit Morphable Model for Localized Head Modeling
-
2022 MimicME: A Large Scale Diverse 4D Database for Facial Expression Analysis
-
2022 Convolutional Mesh Autoencoders for the 3-Dimensional Identification of FGFR-related Craniosynostosis
-
2020 Synthesizing Coupled 3D Face Modalities by Trunk-Branch Generative Adversarial Networks