Konstantinos Vardis
Computer Scientist
My work focuses on computer graphics and vision R&D, with emphasis on rendering, neural graphics, and learned 3D content generation. I hold a PhD from the Athens University of Economics and Business (AUEB) and am currently based in Zürich, Switzerland, working as a Principal Researcher at Huawei's Zürich Research Center.
Research Interests: 3D Generative AI, Light Transport, GPU Computing
Selected Publications (Full list on Google Scholar)
2026. Real-time Rendering with a Neural Irradiance Volume
Eurographics 2026 (accepted)
2026. DreamAnywhere: Object-Centric Panoramic 3D Scene Generation
IEEE/CVF Winter Conference on Applications of Computer Vision (WACV 2026) (Oral)
2023. A Neural Builder for Spatial Subdivision Hierarchies
The Visual Computer (Proceedings of Computer Graphics International 2023)
2020. A Survey of Multifragment Rendering
Computer Graphics Forum – presented at Eurographics 2020
2016. DIRT: Deferred Image-based Ray Tracing
Eurographics/ACM SIGGRAPH Symposium on High Performance Graphics (HPG 2016)
2014. Real-time Radiance Caching using Chrominance Compression
Journal of Computer Graphics Techniques (JCGT) – presented at i3D 2015
2013. Multi-view Ambient Occlusion with Importance Sampling
Proc. of the ACM SIGGRAPH Symposium on Interactive 3D Graphics and Games (i3D 2013)
PhD Thesis
2016. Efficient Illumination Algorithms for Global Illumination in Interactive and Real-Time Rendering
Department of Informatics, Athens University of Economics & Business.
Advisor: Prof. Georgios Papaioannou
Professional Experience (See full CV)
Principal Researcher / Team Lead
Research and development in computer graphics, vision and machine learning at the Cloud Graphics Research Group.
- Responsible for leading team(s) of research scientists and engineers.
- Research and development in 3D Generative AI, Neural Graphics and Neural Rendering.
Postdoctoral Researcher
Research and development on interactive rendering and illumination algorithms.
- Lumibricks (2020–2021): Modular Illumination Transfer for Photorealistic Visualization on Commodity Hardware.
15-month research program co-funded by the European Union and Greek National Funds. - Interactive Photorealistic Image Synthesis (2020): 12-month research program funded by the AUEB Research Center.
- Proof-of-concept implementation of coarse shading technologies for the ARM Mali-G76 Bifrost architecture (2019).
PhD Researcher
- Nomitech Ltd (2016–2017): Visualization Engine for the CostOS software suite.
- PRESIOUS (2013–2016): Predictive digitization, restoration and degradation assessment of cultural heritage objects.
3-year EU-funded STREP (GID: 600533).
- GLIDE (2014–2015): Goal-driven Lighting for Dynamic 3D Environments.
18-month project co-funded by the General Secretariat of Research and Technology and the European Union (ARISTEIA II programme).
Satellite imagery processing: Machine Learning pipelines for visible, infrared, and radar data.
Senior Software Engineer
Think Silicon - Applied Materials
- Development of drivers and software tools for ultra-low power GPUs (Nema), tested on software reference mode and Xilinx FPGA SoCs.
- Design and development of GLOVE, an open-source intermediate layer between OpenGL ES and Vulkan. Achieved EGL 92% and GLES2 98% conformance.
- GPU-WEAR: Ultra-low power heterogeneous GPUs for Wearable/IoT devices. EU-funded H2020 project (GID: 717850), 2016–2018.
AAA game engine rendering: vegetation systems, naval battles, buildings, and UI. Empire: Total War (BAFTA Award).
Academic
Service
Reviewer and IPC member for:
Education
Teaching
Athens University of Economics and Business , Department of Informatics
Teaching Assistant 2012 – 2015
Computer Graphics BSc Course Databases BSc Course
Coordinating lab sessions for over 300 BSc students each year, including course and exam material preparation, teaching, and examining.
Research Software
A multi-purpose platform for virtual operations on digitized cultural heritage objects. Concentrated on object reassembly tasks and measurements in a fully three-dimensional environment.
A scene graph-based deferred rendering API and research platform for testing real-time algorithms for shading and illumination effects using OpenGL. Provided as DLL with ANSI C API. Property of the AUEB Graphics Group, under MIT license.
A mini, CPU-based, tiled ray tracing engine consisting of two files with minimal dependencies. Created for educational purposes. Suitable for rapid prototyping and integration into larger systems.
Reference implementations of common C++ data structures and basic searching/sorting algorithms written in C++11. Intended as a helper on common implementations, not as a performance-optimized library.
An online platform for ray tracing shader development and experimentation. Real-time interactive environment for learning and prototyping ray tracing algorithms.