Kenny Erleben
   

Kenny Erleben

Professor, Head of IMAGE research section

 https://orcid.org/0000-0001-6808-4747

  • Universitetsparken 1

    2100 København Ø

  • Source: Scopus
20022024

Research activity per year

Personal profile

Primary fields of research

My research is about how to build simulators and use data-driven approaches together with simulation. Hence, my work ranges from inventing new numerical methods for simulation to using simulation to generate data for machine learning or learning physical models from real-world data.

Teaching

Computer simulation, computer graphics and computer animation.

CV

Contact Info 

Kenny Erleben 

Department of Computer Science 

University of Copenhagen 

Universitetsparken 5, 

DK-2100, Copenhagen 

Phone: +45 29631108  

Work email: [email protected] 

Personal web page: https://erleben.github.io/ 

Research Group: https://di.ku.dk/english/research/image/ 

ORCID: https://orcid.org/0000-0001-6808-4747 

Google Scholar: https://scholar.google.com/citations?user=CQkvlpUAAAAJ&hl 

 

About me 

I am leading a research section named IMAGE at the University of Copenhagen. IMAGE has about 10 faculty staff, 2 Postdocs, and 20 PhDs. IMAGE is organized into smaller research teams; these cover research areas of computer vision, robotics, medical image analysis, numerical methods, geometry, machine learning, and computer graphics. My team has about 10 people and we are interested in how to build simulators and how to use data-driven approaches together with simulation. Hence, our work ranges from inventing new numerical methods for simulation to using simulation for generating data for machine learning or learning physical models from real-world data. 

 

Research Topics 

Physics-based modeling and simulation, Robotics Automation, Data-driven modeling, Digital Twins. 

 

Recent Academic Positions 

2022-now 

Representing the University of Copenhagen in Manufacturing Academy Denmark (MADE). 

2022-now 

Professor at the Department of Computer Science, University of Copenhagen. 

2021-now 

Head of Section, IMAGE, Department of Computer Science, University of Copenhagen. 

2019-2020 

Scientific Advisor for Alexandra Institute. 

2018-2019 

Part-time Team Lead at Alexandra Institute. 

2012-2015 

Deputy Head at the Department of Computer Science, University of Copenhagen (management period with 3 years reduced research activity). 

2009-2022 

Associate Professor at the Department of Computer Science, University of Copenhagen. 

2005-2009 

Assistant Professor at the Department of Computer Science, University of Copenhagen. 

2005 

Doctor of Philosophy (PhD) in Computer Science, University of Copenhagen. 

2003-now 

Co-creator and continued member of the work group of the Danish Academy of Digital Interactive Entertainment (DADIU). 

 

Selected Major Funding 

2023 

HORIZON-RIA, Intelligent Robotic Endoscopes for Improved Healthcare Services, Coordinator & Principal Investigator (6.189.082 EUR). 

2021 

ErhvervsPhD Danmarks Innovationsfond: Home monitoring of Parkinson’s Disease gait symptoms based on AI and smart insoles (principal investigator, 2.130.000 DKK, collaborator ZenzeTech). 

2020 

Data+ pool (UCPH Strategy 2023 funds): Virtual Kidney (co-applicant, 2.059.836 DKK with Prof Olga Sosnovtseva from the Department of Biomedical Sciences). 

2019 

DFF-Research Project 1: Reward Modelling for Soft Robotics AI (principal investigator, 2.582.150 DKK). 

2019 

ErhvervsPhD Danmarks Innovationsfond: Auto-animation of digital humans using AI methods (ADAM) (principal investigator, 2.130.000 DKK, collaborator Rokoko). 

2018 

TALENT Doctoral Fellowship: Digitalization of the Dental Industry through Deep Learning (principal investigator, 1.600.000 DKK, collaborator 3Shape ApS). 

2018 

H2020 MSCA-IF: EMISSR; Efficient Multibody Interactive Simulation for Soft Robotics, Coordinator (200.000 EUR approx. 1.488.686 DKK).  

2017 

H2020-MSCA-ITN-2017-764644-RAINBOW, Coordinator, and main applicant (4.013.444.88 EUR approx. 29.873.795 DKK). 

 

Brief Summary of Recent Scientific Production 

Total of 129 works. 19 First authored works, 45 last-authored works, 13 single-authored works, and 4 prizes received. Corresponding author on 81 works and 111 peer-reviewed works. According to Google Scholar: 2182 all-time citations, h-idx 23, and four best paper prizes. 

 

Supervision, Advising & Mentoring of Younger Researchers 

I have been supervising 15 PhDs and 8 Postdocs. Mentoring one tenure track assistant professor. I have supervised more than 170 MSc/BSc theses. 

 

Selected Invited Talks within The Last 5 Years 

2023 

Modeling Friction in Interactive Digital Simulations. Talk at Unity Physics Club 

2023 

“Digitale tvillinger vinder frem” (Danish titel). Interview by Ingeniøren 

2022 

Digital Twins for Digital Design and Quality Assessment in Production. Session at Digital Tech Summit 

2022 

Differential simulators, digital twins & data (d3). Talk at International Workshop by DIREC – Digital Research Centre Denmark at University of Southern Denmark 

2022 

Interviewed about simulation and artificial intelligence. DR Podcasts Spøgelset i Maskinen 

2022 

How to most easily tell a robot what to do? Talk at Digital Tech Summit 

2020 

Invited talk at the conference SCA on the Matchstick Model. 

2019 

Presented our research results on data-driven control of soft robots at the Danish IP Fair. 

2019 

Presented RAINBOW results at the EU REA Artificial Intelligence Cluster event in Brussels, Belgium. 

2018 

Invited presentation at SIGGRAPH conference on local mesh-based method for contact point generation. 

2018 

Invited talk at The Centre for Intelligent Machines (CIM), McGill University Montreal, Canada on the topic of iterative methods for contact force solving. 

 

Community Services and Leadership 

I have served regularly as reviewer for international conferences and journals such as SIGGRAPH, Eurographics, IEEE Transactions on Vision and Computer Graphics (TVCG), ACM Transactions on Graphics (TOG), Computers and Graphics (CAG), Computer Graphics Forum (CGF), International Conference on Robotics and Automation (ICRA), IEEE Robotics and Automation Letters (RA-L) and more, during which I have reviewed more than 300 scientific papers. 

 

Selection of 10 Impact Full Publications 

[1] Differentiable Depth for Real2Sim Calibration of Soft Body Simulations, by Kasra Arnavaz, Max Kragballe Nielsen, Paul G. Kry, Miles Macklin, and Kenny Erleben. Computer Graphics Forum, Pages 1-14 (2022). 

Impact: The concept of end-to-end differentiable pipelines is applied to real-world soft robotics using depth sensor technology. 

[2] Contact and friction simulation for computer graphics, by Sheldon Andrews, Kenny Erleben, and Zachary Ferguson. SIGGRAPH '22: ACM SIGGRAPH 2022 Courses, Number 3, Pages 1–172 (2022). 

Impact: Most updated references on work in the field, cover soft body contact bottlenecks in detail. 

[3] Coupling Friction with Visual Appearance, by Sheldon Andrews, Loic Nassif, Kenny Erleben, and Paul G. Kry. Proceedings of the ACM on Computer Graphics and Interactive Techniques, Volume 4, Issue 3, Number 31, Pages 1–20 (2021). 

Impact: New physical model that extends friction modeling to achieve higher fidelity. 

[4] gradSim: Differentiable simulation for system identification and visuomotor control, by Krishna Murthy Jatavallabhula, Miles Macklin, Florian Golemo, Vikram Voleti, Linda Petrini, Martin Weiss, Breandan Considine, Jerome Parent-Levesque, Kevin Xie, Kenny Erleben, Liam Paull, Florian Shkurti, Derek Nowrouzezahrai and Sanja Fidler. International Conference on Learning Representations (ICLR) (2021). 

Impact: The first end-to-end differentiable pipeline with rendering, soft body simulation, and contact. 

[5] Primal/Dual Descent Methods for Dynamics, by Miles Macklin, Kenny Erleben, Matthias Müller, Nuttapong Chentanez, Stefan Jeschke and Tae-Yong Kim. Computer Graphics Forum, Volume 39, Pages 89-100 (2020). 

Impact: A new numerical method for NVIDIA warp that supports differentiable rigid body simulation. 

[6] Local Optimization for Robust Signed Distance Field Collision, by Miles Macklin, Kenny Erleben, Matthias Müller, Nuttapong Chentanez, Stefan Jeschke and Zach Corse. Proceedings of the ACM on Computer Graphics and Interactive Techniques, Volume 3, Issue 1, Number 8, Pages 1–17 (2020). 

Impact: A fast collision detection method for complex geometry used by NVIDIA Isaac, PhysX, MuJuCo, and many more simulators. 

[7] The Matchstick Model for Anisotropic Friction Cones, by Kenny Erleben, Miles Macklin, Sheldon Andrews, and Paul G. Kry. Computer Graphics Forum, Volume 39, Pages 450-461 (2019). 

Impact: Modeling for higher fidelity leads to new models. 

[8] Non-Smooth Newton Methods for Deformable Multi-Body Dynamics, by Miles Macklin, Kenny Erleben, Matthias Müller, Nuttapong Chentanez, Stefan Jeschke and Viktor Makoviychuk. ACM Transactions on Graphics, Volume 38, Issue 5O, Number 140, Pages 1–20 (2019). 

Impact: The paper behind the numerical method behind NVIDIA Isaac. 

[9] Methodology for Assessing Mesh-Based Contact Point Methods, by Kenny Erleben. ACM Transactions on Graphics, Volume 37, Issue 3, Number 39, Pages 1–30 (2018). 

Impact: The standard for testing contact point generation for simulation. 

[10] Rigid Body Contact Problems using Proximal Operators, by Kenny Erleben. SCA '17: Proceedings of the ACM SIGGRAPH / Eurographics Symposium on Computer Animation, Issue 13, Pages 1–12 (2017). 

Impact: The most generic form of projected gauss-seidel type of solvers, an industry standard for contact handling. 

 

 

Current research

See https://erleben.github.io/ for details

Keywords

  • Faculty of Science
  • Computational Physics
  • Applied Math
  • Scientific Computing
  • Physics based Animation
  • Numerical Optimization
  • Constrainted Optimization
  • Nonlinear Nonsmooth Optimization Problems
  • Complementarity Problems
  • Collision Detection
  • Surgical Simulators
  • Computer Games Physics
  • Computer Science

Collaborations and top research areas from the last five years

Recent external collaboration on country/territory level. Dive into details by clicking on the dots or
Explore network further
View all 144 research outputs