Stefan Stevšić

PhD Student
stefan.stevsic@inf.ethz.ch
ETH Zurich, Department of Computer Science, STD, H 29.1
Universitätstrasse 6, 8092 Zürich, Switzerland

Biography

I am a final year Ph.D. student at the Department of Computer Science at ETH Zurich, working under the supervision of Prof. Dr. Otmar Hilliges. My focus lies in the areas of Computer Vision and Control Policy Learning, with the goal of enabling robots to learn new tasks from human demonstrations. I am interested in 6D pose estimation of rigid objects from RGB-D data, which allows robots to learn new manipulation tasks. For learning motions skills from human demonstration, I am interested in Imitation Learning and Reinforcement Learning approaches.


Additional Links:

Publications

2021

Improved Learning of Robot Manipulation Tasks via Tactile Intrinsic Motivation
Nikola Vulin, Sammy Christen, Stefan Stevšić, Otmar Hilliges
IEEE Robotics and Automation Letters (Volume: 6, Issue: 2), 2021
Video Paper Code

2020

Spatial Attention Improves Iterative 6D Object Pose Estimation
Stefan Stevšić, Otmar Hilliges
2020 International Conference on 3D Vision (3DV), 2020
Video Paper Code
Capturing Subjective First-Person View Shots with Drones for Automated Cinematography
Amirsaman Ashtari, Stefan Stevšić, Tobias Nägeli, Jean-Charles Bazin, Otmar Hilliges
ACM Transactions on Graphics (Proceedings of ACM SIGGRAPH) (Volume: 39, Issue: 5), 2020
Video Paper
Learning to Assemble: Estimating 6D Poses for Robotic Object-Object Manipulation
Stefan Stevšić, Sammy Christen, Otmar Hilliges
IEEE Robotics and Automation Letters (Volume: 5, Issue: 2), 2020
Video Paper Data Code

2019

Demonstration-Guided Deep Reinforcement Learning of Control Policies for Dexterous Human-Robot Interaction
Sammy Christen, Stefan Stevšić, Otmar Hilliges
International Conference on Robotics and Automation (ICRA), 2019
Video Paper Code

2018

Sample Efficient Learning of Path Following and Obstacle Avoidance Behavior for Quadrotors
Stefan Stevšić, Tobias Nägeli, Javier Alonso-Mora, Otmar Hilliges
IEEE Robotics and Automation Letters (Volume: 3, Issue: 4), 2018
Video Paper
Optimizing for Aesthetically Pleasing Quadrotor Camera Motion
Christoph Gebhardt, Stefan Stevšić, Otmar Hilliges
ACM Transactions on Graphics (Proceedings of ACM SIGGRAPH) (Volume: 37, Issue: 4), 2018
Video Paper

2016

Airways: Optimization-based Planning of Quadrotor Trajectories according to High-Level User Goals
Christoph Gebhardt*, Benjamin Hepp*, Tobias Nägeli, Stefan Stevšić, Otmar Hilliges * Equal contribution
SIGCHI Conference on Human Factors in Computing Systems, 2016
Video Paper

Teaching

Student Theses

2020
MA Nikola Vulin Learning to Control Contact Forces in Robotic Manipulation
2018
MA Sammy Christen Demonstration-Guided Deep Reinforcement Learning of Control Policies for Human-Robot Interaction
2018
MA Lyuan Hsu Transfer Learning for Mapless Quadrotor Navigation Using Recurrent Neural Network
2018
SA Siwei Zhang One-Shot Imitation Learning for Particle Reaching with Multiple Targets

Courses