Zhiyuan Hu

Master Thesis

2023

Exploring Safety for Reinforcement Learning in Autonomous Drone Racing

Supervised by Angel Romero and Andrea Carron

Tim Flueckiger

Semester Thesis

2023

Developing Smart Vision Assistive Technologies

Supervised by Giovanni Cioffi

Mertcan Akcay

Master Thesis

2023

Learning to Fly by Watching Youtube Videos

Supervised by Marco Cannici and Angel Romero

Abhiram Shenoi

Semester Thesis

2023

Data-Driven Generation of Synthetic Events

Supervised by Nico Messikommer and Mathias Gehrig

Michele Pagani

Master Thesis

2023

Exploring the Viability of Autonomous Drone Flight Using an Off-Board Controller

Supervised by Angel Romero and Leonard Bauersfeld

Gokul Raju Govinda Raju

Semester Thesis

2023

Pushing the Limits of Optical Flow Estimation for Event Cameras

Supervised by Nikola Zubic and Marco Cannici

Marwan Ibrahim

Master Thesis

2023

Event-Guided Depth Sensing based on Spatially Asynchronous Scanning

Supervised by Manasi Muglikar and Pierre Giraud

Marco Magno

Master Thesis

2023

Benchmarking Visual Localization Methods for Drone Racing

Supervised by Drew Hanover and Giovanni Cioffi

Ivan Alberico

Master Thesis

2023

Vision-Based Navigation Initialization for Mid-Air Helicopter Delivery on Mars

Thesis at NASA JPL. RPG supervisor Giovanni Cioffi

Shreedhar Govil

Master Thesis

2023

Spiking Neural Networks for Drone Navigation

Supervised by Stanislaw Wozniak (IBM) and Mathias Gehrig

Caterina Caccavella

Master Thesis

2023

Low-power event-based face detection with asynchronous neuromorphic hardware and spiking neural networks

Supervised by Lyes Khacef (Sony), Federico Paredes-Valles (Sony), and Marco Cannici

Adarsh Patnaik

Semester Thesis

2023

Model-based Reinforcement Learning for Agile Drone Control

Supervised by Angel Romero

Victoria Catalan Pastor

Semester Thesis

2023

Developing Smart Vision Assistive Technologies

Supervised by Giovanni Cioffi

Roberto Pellerito

Master Thesis

2023

End-To-End Learned VO with Event Cameras for Moon Landing

Supervised by Marco Cannici, Daniel Gehrig, Joris Belhadj (ESA), Olivier Dubois-Matra (ESA) and Massimo Casasco (ESA)

Agnar Petursson and Levente Foeldesi

Master Project

2023

Weakly Supervised Gate Detection for Drone Racing

Supervised by Christian Peiffer, Daniel Gehrig and Leonard Bauersfeld

Chunwei Xing

Master Thesis

2023

Vision-Based Offboard Flight

Supervised by Leonard Bauersfeld, Yunlong Song and Jiaxu Xing

Simone Nascivera

Semester Project

2023

Multirotor Extension of Agilicious

Supervised by Leonard Bauersfeld and Angel Romero

Narek Bayanduryan

Master Thesis

2023

Improvement of Quadrotor State Estimation with Force Sensors

Supervised by Angel Romero, Drew Hanover and Giovanni Cioffi

Siva Chidambaram

Semester Thesis

2023

Siva worked on High-Frequency Scanning and Depth Estimation in Event Based Structured Light System.

Supervised by Manasi Muglikar and Diederik Moeys

Mengqi Wang

Master Thesis

2023

Mengqi worked on Neural Implicit Surface Reconstruction for Reflective Surfaces.

Supervised by Marco Cannici, Dal Noguer (Magic Leap), and David Ferstl (Magic Leap)

Yifei Liu

Master Thesis

2023

Yifei worked on sparse computation with Vision Transformers.

Supervised by Mathias Gehrig and Nico Messikommer

Asude Aydin

Master Thesis

2023

Asude worked on hybrid ANN-SNN architectures.

Supervised by Mathias Gehrig and Daniel Gehrig

Julia Prozorova

Independent Study

2023

Julia Prozorova worked on learning a velocity controller for drones.

Supervised by Leonard Bauersfeld

Rong Zou

Semester Thesis

2023

Rong Zou worked on image reconstruction with event cameras

Supervised by Manasi Muglikar and Nico Messikommer

Kaixi Bao

Semester Project

2023

Kaixi Bao applied Confidence Region Bayesian Optimization (CRBO) to the problem of tuning a Model Predictive Contouring Control strategy for drone racing.

Supervised by Angel Romero and Andrea Carron

Pietro Bonazzi

Master Project

2023

Pietro worked on point cloud reconstruction and denoising via learned-based rendering features

Supervised by Marco Cannici and Marie-Julie Rakotosaona (Google)

Jonas Matt

Semester Project

2023

Jonas integrated human drone racing data into a multi-agent reinforcement learning pipeline and studied the resulting policy behavior

Supervised by Drew Hanover and Yunlong Song

Iuliia Prozorova and Adriana Mohap

Master Project

2023

Iuliia and Adriana worked on Learning 3D Reconstruction for an Event-based Structured Light System

Supervised by Manasi Muglikar and Daniel Gehrig

Zhaoning Sun

Master Thesis

2022

Zhaoning worked on End-to-End Learned Vision-Based Navigation for Quadcopters.

Supervised by Leonard Bauersfeld and Elia Kaufmann

Rupal Saxena and Shivam Adarsh

Master Project

2022

Rupal and Shivam investigated how gravity alterations affect self-motion estimation in human drone pilots.

Supervised by Christian Pfeiffer and Manasi Muglikar

Carter Fang

Master Thesis

2022

Carter developed a data-driven keypoint tracker for event cameras.

Supervised by Nico Messikommer and Mathias Gehrig

Gonca Yilmaz and Shreedhar Govil

Master Project

2022

Gonca and Shreedhar worked on policy search for model predictive contouring control tunning.

Supervised by Angel Romero and Yunlong Song

Kexin Shi

Independent Study

2022

Kexin worked on learning perception-aware agile flight in cluttered environment.

Supervised by Yunlong Song and Dr. Robert Penicka

Jiawei Fu

Internship

2022

Jiawei worked on learning deep sensorimotor policies for vision-based autonomous drone racing.

Supervised by Yunlong Song

Nicola Carrino

Master Thesis

2022

Nicola worked on event-based motion segmentation.

Supervised by Mathias Gehrig

Yang Miao

Semester Thesis

2022

Yang developed a feature tracking method for a novel type of event camera.

Supervised by Nico Messikommer and Daniel Gehrig

Florian Trautweiler

Master Thesis

2022

Florian worked on reinforcement learning for super-human performance in autonomous drone racing.

Supervised by Christian Pfeiffer and Yunlong Song

Michele Graziano

Master Thesis

2022

Michele worked on Hardware and Software Codesign for Visual-Inertial Odometry.

Supervised by Angel Romero, Drew Hanover and Giovanni Cioffi

Matteo Visona

Master Project

2022

Matteo worked on improving the efficiency of graph neural networks.

Supervised by Nico Messikommer and Daniel Gehrig

Nico Schulthess

Semester Project

2022

Nico worked on learning-based simulation of synthetic events from video.

Supervised by Daniel Gehrig and Mathias Gehrig

Konstantin Kalenberg

Semester Project

2022

Konstantin worked on accelerating EKLT in EKLT-VIO.

Supervised by Daniel Gehrig and Mathias Gehrig

Pietro Bonazzi and Mengqi Wang

Master Project

2022

Mengqi and Pietro worked on 3D scene generation using scene graphs.

Supervised by Nico Messikommer and Yunlong Song

Cafer Mertcan Akcay

Semester Thesis

2022

Mert worked on using reinforcement learning for offboard control of a racing drone.

Supervised by Christian Pfeiffer, Angel Romero, and Yunlong Song

David Oort Alonso

Semester Thesis

2022

David worked on collaborative aerial transport using deep reinforcement learning.

Supervised by Yunlong Song and Angel Romero

Mirlan Karimov

Semester Thesis

2022

Mirlan worked on creating an optical flow simulation environment

Supervised by Mathias Gehrig and Manasi Muglikar

Xiao'ao Song

Master Thesis

2022

Xiao'ao Song worked on combining stereo and lidar for dense depth estimation.

Supervised by Mathias Gehrig and Nico Messikommer

Kexin Shi and Yifei Liu

Master Project

2022

Kexin and Yifei worked efficient spatio-temporal processing of event data.

Supervised by Mathias Gehrig and Nico Messikommer

Ivan Alberico

Semester Thesis

2021

Ivan worked on learning event generation.

Supervised by Daniel Gehrig and Mathias Gehrig

Zhaoning Sun

Semester Thesis

2021

Zhaoning worked on transferring a segmentation network from frames to events.

Supervised by Nico Messikommer and Daniel Gehrig

Ulrich Steger

Semester Thesis

2021

Ulrich worked on a motion-tracking camera rig.

Supervised by Leonard Bauersfeld and Christian Pfeiffer

Andreas Voigt

Semester Thesis

2021

Andreas worked on learning 3D reconstruction with stereo event camera.

Supervised by Manasi Muglikar and Nico Messikommer

Aurora Colagiorgio

Master Thesis

2021

Aurora worked on learning IMU odometry.

Supervised by Giovanni Cioffi, Christian Pfeiffer and Leonard Bauersfeld

Florian Trautweiler

Semester Thesis

2021

Florian worked on Benchmarking Algorithms for Autonomous Drone Racing.

Supervised by Christian Pfeiffer and Leonard Bauersfeld

Florian Mahlknecht

Internship

2021

Florian worked on event-based VIO algorithms for Mars-like environments in collaboration with NASA JPL.

Supervised by Daniel Gehrig and Mathias Gehrig

Friedrich Rockenbauer

Internship

2021

Friedrich worked on VIO data collection including event cameras for Mars-like environments in collaboration with NASA JPL.

Supervised by Mathias Gehrig and Daniel Gehrig

Fang Nan

Master Thesis

2021

Fang worked on model-predictive control under rotor failures.

Supervised by Philipp Foehn and Sihao Sun

Michael Schmid

Semester Thesis

2021

Michael worked on MPC control for power line inspection.

Supervised by Giovanni Cioffi and Javier Hidalgo

Shashank Singh

Master Thesis

2021

Shashank worked on optical and scene flow integration in the CARLA simulator.

Supervised by Mathias Gehrig and Javier Hidalgo

Julius Erbach

Master Thesis

2021

Julius worked on stereo event to RGB alignment.

Supervised by Daniel Gehrig and Nico Messikommer

Giovanni Lippolis

Master Thesis

2021

Giovanni worked on event-based video stabilization.

Supervised by Javier Hidalgo

Beat Hubmann

Semester Thesis

2021

Beat worked on Lidar-based mapping

Supervised by Mathias Gehrig and Daniel Gehrig

Qi Ma

Semester Thesis

2021

Qi worked on calibration of event-based active depth sensor.

Supervised by Manasi Muglikar and Mathias Gehrig

Xiao'ao Song

Semester Thesis

2021

Xiao'ao worked on calibration of event-based active depth sensor.

Supervised by Manasi Muglikar and Mathias Gehrig

Selim Naji

Master Thesis

2021

Selim worked on Stepwise Regression MPC for agile quadrotor control.

Supervised by Sihao Sun and Elia Kaufmann

Silvan Loew

Semester Thesis

2021

Silvan worked on online trajectory replanning using Gaussian Processes.

Supervised by Giovanni Cioffi and Robert Penicka

Frederic Letsch

Semester Thesis

2021

Frederic worked on line aware MPC for power line inspection.

Supervised by Giovanni Cioffi and Javier Hidalgo

Friedrich Rockenbauer

Semester Thesis

2021

Generation of a cross compilation pipeline for cutting edge visual-inertial odometry.

Supervised by Javier Hidalgo and Thomas Längle

Antonia Mosberger

Semester Thesis

2021

Antonia worked on dense reconstruction with active event-based depth sensing.

Supervised by Manasi Muglikar and Javier Hidalgo

Kira Erb

Semester Thesis

2021

Kira worked on benchmarking active event-based depth algorithms.

Supervised by Manasi Muglikar and Javier Hidalgo

Nando Kaeslin

Semester Thesis

2021

Nando investigated the effects of camera field of view on first-person view drone racing.

Supervised by Christian Pfeiffer and Yunlong Song

Nicola Carrino

Semester Thesis

2021

Nicola worked on combining Lidar and cameras for depth estimation.

Supervised by Mathias Gehrig and Nico Messikommer

Simone Arreghini

Master Thesis

2021

Simone worked on learning vision-based RL policies for dynamic obstacle avoidance.

Supervised by Antonio Loquercio and Yunlong Song

Lorenzo Ferrini

Master Thesis

2021

Lorenzo worked on sample-efficient reinforcement learning from images.

Supervised by Yunlong Song and Antonio Loquercio

Mario Bonsembiante

Master Thesis

2021

Mario worked on training an universal controller for drone flight with RL.

Supervised by Antonio Loquercio and Elia Kaufmann

Yilun Wu

Master Thesis

2021

Yilun worked on event- and frame-based frame interpolation from unaligned inputs.

Supervised by Daniel Gehrig and Mathias Gehrig

Mario Millhaeusler

Master Thesis

2021

Mario worked on combining frame and event camera data for dense stereo matching

Supervised by Mathias Gehrig and Daniel Gehrig

Simon Wengeler

Master Thesis

2021

Simon worked on Human Attention for End-to-End Drone Racing.

Supervised by Christian Pfeiffer and Antonio Loquercio

Mats Steinweg

Master Thesis

2021

Mats worked on Planning Minimum-time Trajectory for Drone Racing Using Reinforcement Learning.

Supervised by Yunlong Song and Elia Kaufmann

Leonard Bauersfeld

Master Thesis

2021

Leonard worked on system identification for quadrotors using a hybrid approach that combines blade-element-momentum theory with a learning-based correction.

Supervised by Elia Kaufmann and Philipp Foehn

Alexander Dietsche

Semester Thesis

2021

Alexander Dietsche developed an event-based line tracker for power line inspection.

Supervised by Giovanni Cioffi and Javier Hidalgo

Jiaxu Xing

Semester Thesis

2021

Jiaxu Xing worked on flight trajectory modeling for human-piloted drone racing.

Supervised by Christian Pfeiffer and Elia Kaufmann

Amir Hadzic

Semester Thesis

2021

Amir Hadzic worked on non-causal, continuous time intensity estimation from events.

Supervised by Daniel Gehrig and Mathias Gehrig

Florian Mahlknecht

Semester Thesis

2021

Florian worked on event-based stereo vision for scene flow understanding in autonomous driving

Supervised by Mathias Gehrig and Daniel Gehrig

Gian Erni

Bachelor Thesis

2021

Gian worked on Eventmare: from event camera simulator to egomotion estimation.

Supervised by Javier Hidalgo and Yunlong Song

Mike Suter

Master Project

2020

Mike developed a platform for multi-camera SLAM.

Supervised by Manasi Muglikar and Javier Hidalgo

Thomas Huber

Master Project

2020

Thomas developed a platform for multi-camera SLAM.

Supervised by Manasi Muglikar and Javier Hidalgo

Alessandro Saviolo

Master Thesis

2020

Alessandro developed a framework for automatic controller tuning.

Supervised by Antonio Loquercio, Philipp Foehn, and Elia Kaufmann

Willem Aarents

Master Thesis

2020

Willem worked on creating a sensor setup for collection of an event-based vision driving dataset.

Supervised by Mathias Gehrig and Daniel Gehrig

Dingguang Jin

Master Thesis

2020

Dingguang worked on generating High resolution video from events.

Supervised by Daniel Gehrig and Mathias Gehrig

Selim Naji

Semester Thesis

2020

Selim worked on Bridging the gap between Gazebo and Unity3D for quadrotor simulation.

Supervised by Yunlong Song, Elia Kaufmann, and Antonio Loquercio

HaoChih Lin

Master Thesis

2020

HaoChih worked on reinforcement learning for race car overtaking in Grand Turismo Sport.

Supervised by Yunlong Song and Elia Kaufmann

Guillem Torrente

Master Thesis

2020

Guillem worked on designing a model predictive controller for quadrotors that is augmented with Gaussian Processes.

Supervised by Elia Kaufmann and Philipp Foehn

Alex Schindler

Master Project

2020

Alex worked on ground segmentation for autonomous drone landing.

Supervised by Kunal Shrivastava, Kevin Kleber and Antonio Loquercio

Roger Staubli

Master Project

2020

Roger worked on ground segmentation for autonomous drone landing.

Supervised by Kunal Shrivastava, Kevin Kleber and Antonio Loquercio

Simon Frischknecht

Master Project

2020

Simon worked on ground segmentation for autonomous drone landing.

Supervised by Kunal Shrivastava, Kevin Kleber and Antonio Loquercio

Michelle Rueegg

Master Thesis

2020

Michelle worked on combining frames and events for asynchronous multi-modal monocular depth prediction.

Supervised by Daniel Gehrig and Mathias Gehrig

Patrick Barton

Semester Thesis

2020

Patrick worked on combining frames and events for high-speed video.

Supervised by Mathias Gehrig and Daniel Gehrig

Maxime Choulika

Internship

2020

Maxime worked on learned data-efficient visual localization.

Supervised by Titus Cieslewski

Yumin Sun

Master Thesis

2020

Yumin worked on event-based semantic segmentation.

Supervised by Mathias Gehrig and Daniel Gehrig

Yaswanth Mummaneni

Semester Thesis

2020

Yaswanth worked on analyzing event-based VIO for agile flight.

Supervised by Elia Kaufmann and Mathias Gehrig

Daniel Mouritzen

Master Thesis

2020

Daniel worked on the problem of autonomous navigation with deep-RL methods.

Supervised by Antonio Loquercio and Mathias Gehrig

Amedeo Fabris

Semester Thesis

2020

Amedeo worked on the characterization of the impact of the morphology of a quadrotor on its thrust capabilities.

Supervised by Davide Falanga and Kevin Kleber

Christoph Meyer

Semester Thesis

2020

Christoph worked on the autonomous control of morphology for the foldable drone.

Supervised by Antonio Loquercio and Elia Kaufmann

Mario Millhaeusler

Semester Thesis

2020

Mario worked on creating a multi-sensor setup for autonomous car dataset collection.

Supervised by Mathias Gehrig and Daniel Gehrig

Florian Fuchs

Master Thesis

2019

Florian worked on reinforcement learning for car racing in Grand Turismo Sport.

Supervised by Elia Kaufmann and Yunlong Song

Nico Messikommer

Master Thesis

2019

Nico worked on self-supervised interest point selection.

Supervised by Elia Kaufmann and Titus Cieslewski

Harshit Khurana

Master Thesis

2019

Harshit worked on Model Predictive Control for morphing quadrotors

Supervised by Davide Falanga and Philipp Foehn

Balazs Nagy

Master Thesis

2019

Balazs worked on optimizing a VIO frontend for GPU-targetted deployment. By carefully implementing and profiling feature detection, tracking, and non-maxima suppresion on the GPU, his pipeline reaches 330 fps on a nVidia Jetson TX2.

Supervised by Philipp Foehn

Roy Rutishauser

Master Thesis

2019

Roy trained a neural network to detect Aruco tags under difficult conditions.

Supervised by Titus Cieslewski

Giovanni Cioffi

Master Thesis

2019

Giovanni implemented visual teach and high-speed repeat on a real quadrotor.

Supervised by Titus Cieslewski

Guillem Torrente

Semester Thesis

2019

Guillem explored using machine learning to improve the accuracy of pure inertial odometry.

Supervised by Zichao Zhang and Elia Kaufmann

Livio Giacomini

Semester Thesis

2019

Livio explored learning-based approaches that predict dense depth from RGB and sparse depth data.

Supervised by Elia Kaufmann

Juichung Kuo

Independent Research Project

2019

Juichung extended our visual-inertial pipeline to handle different camera configurations with minimum parameter tuning required.

Supervised by Zichao Zhang

Yasser Belal

Semester Thesis

2019

Yasser explored adaptive control methods for high-performance quadrotor flight.

Supervised by Dario Brescianini

Mattia Segu'

Semester Thesis

2019

Mattia developed a framework for uncertainty estimation in Deep Learning.

Supervised by Antonio Loquercio

Amadeus Oertel

Master Thesis

2019

Amadeus created a CNN-based descriptor for place recognition that fuses appearance and 3D structural information.

Supervised by Titus Cieslewski and Mathias Gehrig

Kevin Kleber

Master Thesis

2019

Kevin worked on low-latency obstacle avoidance using event cameras.

Supervised by Davide Falanga

Rika Sugimoto

Internship

2019

Rika worked on high speed vision-based control with event cameras.

Supervised by Mathias Gehrig and Dario Brescianini

Samuel Bryner

Master Thesis

2018

Samuel developed an algorithm to track the pose of an event camera given a map of the environment. His approach uses both the shape and intensity information of the scene map, as well as the camera motion to accurately predict the observed events produced by the camera.

Supervised by Henri Rebecq, Guillermo Gallego and Mathias Gehrig

Kunal Shrivastava

Master Thesis

2018

Kunal developed a system for precise and safe, vision-based landing of a cargo drone in cluttered environments. This project was done in collaboration with the Laboratory of Intelligent Sysetms at EPFL.

Supervised by Jeff Delmerico

François Elvinger

Semester Thesis

2018

François Elvinger worked on efficiently selecting features in visual odometry to reduce computational cost.

Supervised by Zichao Zhang

Kristin Schläpfer

Semester Thesis

2018

Kristin worked with Roy Rutishauser on real-time detection and localization of windows and doors with a quadrotor.

Supervised by Jeff Delmerico

Roy Rutishauser

Semester Thesis

2018

Roy worked with Kristin Schläpfer on real-time detection and localization of windows and doors with a quadrotor.

Supervised by Jeff Delmerico

Thomas Eppenberger

Semester Thesis

2018

Thomas worked on pruning channels of NetVLAD, a neural network for place recognition.

Supervised by Titus Cieslewski and Mathias Gehrig

Yves Kompis

Bachelor Thesis

2018

Yves worked on a tag + VO based localization system to be used for evaluation of SLAM systems.

Supervised by Titus Cieslewski and Guillermo Gallego

Anna Leidi

Semester Thesis

2018

Anna worked on minimum time trajectory generation for quadrotors.

Supervised by Davide Falanga and Suseong Kim

Moritz Zimmermann

Semester Thesis

2018

Moritz worked on monocular depth estimation from single images using deep learning.

Supervised by Elia Kaufmann and Antonio Loquercio

Simon Muntwiler

Semester Thesis

2018

Simon worked on deep reinforcement learning for drone racing.

Supervised by Antonio Loquercio and Elia Kaufmann

Dan Jia

Master Thesis

2018

Dan worked on continuous structure from motion (SFM) with event cameras. He designed novel SFM methods to jointly estimate the camera motion and the scene depth based on the continuous epipolar constraint.

Supervised by Guillermo Gallego and Henri Rebecq

Andreas Ziegler

Master Thesis

2018

Andreas developed a representation for frontier-based exploration that is robust to state estimate drift and verified it in 2D.

Supervised by Titus Cieslewski

Daniel Gehrig

Master Thesis

2018

Daniel designed an asynchronous feature tracking algorithm that uses both events and frames. His approach uses the full photometric information from the frames to track features in the event stream with a high accuracy and robustness.

Supervised by Henri Rebecq and Guillermo Gallego

Bojana Nenezic

Master Thesis

2018

Bojana worked on learning 6-DOF camera motion with an event camera.

Supervised by Guillermo Gallego, Antonio Loquercio and Henri Rebecq

Raffael Theiler

Bachelor Thesis

2018

Raffael extended our event camera simulator to use photorealistic images from Unreal Engine, and also improved the performance of the simulator.

Supervised by Henri Rebecq

Nils Funk

Semester Thesis

2017

Nils worked on obstacle avoidance for quadrotors using a depth sensor.

Supervised by Davide Falanga and Philipp Foehn

Kunal Shrivastava

Semester Thesis

2017

Kunal implemented an online loop closure detection system and tested the impact of different parameters.

Supervised by Zichao Zhang and Titus Cieslewski

Samuel Bryner

Semester Thesis

2017

Samuel built a hardware synchronized device composed of a mDAVIS 346b and a Bluefox camera, to combine the best of event cameras and standard cameras.

Supervised by Henri Rebecq

Robin Scherrer

Master Thesis

2017

Robin worked on control and planning for quadrotors manipulating slung-loads.

Supervised by Naveen Kuppuswamy, Davide Falanga and Philipp Foehn

Patrick Widmer

Bachelor Thesis

2017

Patrick worked on estimating rotation from a single blurred image using deep learning.

Supervised by Zichao Zhang and Antonio Loquercio

Yawei Ye

Master Project

2017

Yawei worked on Place Recognition from the structural information contained in Semi-Dense Point Clouds.

Supervised by Titus Cieslewski and Antonio Loquercio

Christopher Tchervenkov

Master Project

2017

Chris worked on active exploration and visual saliency mapping with a quadrotor in a search-and-rescue scenario.

Supervised by Jeff Delmerico

Elia Kaufmann

Master Project

2017

Elia worked on obstacle avoidance and rapid exploration with the Intel Realsense RGBD camera.

Supervised by Titus Cieslewski

Philipp Foehn

Master Project

2017

Philipp worked on nonlinear trajectory optimization for quadrotors manipulating cable-suspended payloads.

Supervised by Davide Falanga and Naveen Kuppuswamy

Jihwan Youn

Semester Project

2016

Jihwan implemented a 3D reconstruction algorithm using a moving stereo pair of event cameras.

Supervised by Henri Rebecq and Horstschaefer

Kaju Bubanja

Bachelor Project

2016

Kaju simulated MBZIRC challenge 3 (collaborative object collection) and designed a high-level controller for it.

Supervised by Titus Cieslewski

Nico van Duijn

Semester Project

2016

Nico implemented a system for autonomously picking up and transporting objects with a quadrotor.

Supervised by Jeff Delmerico

Robin Scherrer

Semester Project

2016

Robin focused on the execution of aggressive maneuvers on RPG quadrotors using Drake, Matlab and ROS.

Supervised by Davide Falanga and Naveen Kuppuswamy

Maria Chiara Giorgetti

Semester Project

2016

Maria Chiara worked on the integration between ROS and Drake for the RPG quadrotors.

Supervised by Davide Falanga and Naveen Kuppuswamy

Alessio Zanchettin

Master Project

2016

Alessio investigated strategies for autonomous quadrotor landing on a moving platform using only onboard sensing and computing.

Supervised by Davide Falanga

Micha Brunner

Master Project

2016

Micha Brunner developed an LQR control scheme to transport large slungloads with a quadrotor.

Supervised by Michael Gassner

Timo Horstschäfer

Master Project

2016

Timo developed an algorithm to perform parallel tracking, depth estimation, and image reconstruction with an event camera. For his Thesis, Timo received the 2016 Fritz Kutter Award for Industry Related Thesis in Computer Science.

Supervised by Guillermo Gallego and Elias Mueggler

Filippo Martinoni

Master Project

2016

Filippo developed localization and SLAM for robots equipped with only a couple of terarangers (and an IMU for SLAM).

Supervised by Davide Scaramuzza and Titus Cieslewski

Mathis Kappeler

Master Project

2016

Mathis worked on place recognition and loop closure from sparse feature-based maps.

Supervised by Titus Cieslewski

Philipp Föhn

Semester Project

2016

Philipp focused on the development of an estimator of the external forces acting on a quadrotor and a force-based control approach.

Supervised by Davide Falanga and Michael Gassner

Valentin Wüest

Semester Project

2016

Valentin developed a control scheme to transport a load with a team of two UAVs.

Supervised by Michael Gassner and Davide Falanga

Jonathan Huber

Semester Project

2016

Jonathan designed a localization pipeline that allows a ground robot using a 3D LiDAR sensor to navigate within a dense map vreated from an aerial robot.

Supervised by Jeff Delmerico and Elias Mueggler

Benjamin Imbach

Semester Project

2016

Benjamin developed a novel sensor model to characterize a new lightweight depth camera (Intel RealSense) and deploy it on a quadrotor for performing volumetric mapping.

Supervised by Jeff Delmerico

Yawei Ye

Independent Research Project

2016

Yawei investigated the performance of sparse and dense 3D mapping for the purpose of fast navigation and obstacle avoidance, and developed a sparse mapping software framework based on bounding boxes.

Supervised by Jeff Delmerico

Gianluca Cesari

Internship

2016

Gianluca worked on hardware development related to the participation of RPG in the Mohamed Bin Zayed Internation Robotics Challenge.

Supervised by Michael Gassner and Jeff Delmerico

Julia Nitsch

Internship

2016

Julia developed a system for terrain classification in a search and rescue scenario, to help enable a collaborative air-ground robot team to navigate through a previously unknown environment.

Supervised by Jeff Delmerico, Elias Mueggler, and Christian Forster

Beat Kueng

Master Thesis

2016

Beat designed, implemented, and evaluated an event-based visual-odometry pipeline for the DAVIS. The method is based on features extracted from the frames that are then tracked using the events.

Supervised by Elias Mueggler and Guillermo Gallego

Kevin Egger

Semester Thesis

2016

Kevin developed a fallback-landing routine for situations where the vision-based state estimation on our quadrotors fails. It enables a controlled descent and hence prevents our quadrotors from crashing in case of a failure.

Supervised by Matthias Faessler and Davide Falanga

Mathis Kappeler

Master Project

2016

Mathis implemented and evaluated a variety of auto-exposure algorithms using a lighting testbed that he designed and built. His works increases robustness in autonomous, vision-based flight in challenging lighting conditions.

Supervised by Elias Mueggler and Manuel Werlberger

Raphael Meyer

Master Thesis

2015

Raphael designed a custom quadrotor platform which is becoming the new standard platform for research on vision-based drones at the Robotics and Perception Group. In this project, he designed custom electronics, hardware, and software for these quadrotors.

Supervised by Matthias Faessler and Flavio Fontana

Imanol Studer

Master Project

2015

Imanol compared, implemented, and evaluated different strategies to estimate a person's head orientation. Such algorithms enable quadrotors to remain at a constant position relative to a person.

Supervised by Elias Mueggler

Jon Lund

Master Thesis

2015

Jon worked on pose tracking for event-based sensors. His algorithm allows high-speed camera pose tracking with low latency and high frequency.

Supervised by Guillermo Gallego and Elias Mueggler

Pavel Vechersky

Semester Thesis

2015

Pavel developed a system for performing active dense 3D reconstruction using flying robots. His work extended the lab reconstruction demo to explore and reconstruct the environment without a precomputed set of waypoints to visit.

Supervised by Reza Sabzevari and Jeff Delmerico

Joachim Ott

Semester Thesis

2015

Joachim worked on vision-based surface classification for micro aerial vehicles. His project was a collaboration with IDSIA in Lugano.

Supervised by Christian Forster and Elias Mueggler

Micha Brunner

Semester Thesis

2015

Micha worked on coordinated flight with quadrotors using our infrared pose-estimation system. The incorporation of these relative robot measurements enabled autonomous flights over areas with little or no texture.

Supervised by Elias Mueggler

Igor Bozic

Master Project

2015

Igor worked on a high-frequency position controller for the KUKA youBot arm that allows to follow end-effector trajectories both fast and precisely.

Supervised by Elias Mueggler, Flavio Fontana and Matthias Faessler

Lorenz Wellhausen

Semester Thesis

2015

Lorenz worked on Dense Multi-Body Structure from Motion from a sequence of stereo images. He used Trajectory Flow to find the dense image correspondences and DBSCAN for segmenting the dense trajectories.

Supervised by Reza Sabzevari and Manuel Werlberger

David Tedaldi

Semester Thesis

2015

David worked on feature tracking with the DAVIS. He developed an algorithm to detect image features in the frames and track them in time using an event-based approach.

Supervised by Guillermo Gallego and Elias Mueggler

Michael Gassner

Master Thesis

2015

Michael developed a framework to design optimal trajectories for vision-based quadrotors. These trajectories respect the dynamical constraints of the quadrotors as well as a maximum admissible optical flow of their onboard cameras.

Supervised by Matthias Faessler and Flavio Fontana

Stefan Isler

Semester Thesis

2015

Stefan worked on Active Dense 3D Reconstruction. He developed a pipeline to autonomously select the next camera view in order to have a complete 3D model of a given object.

Supervised by Reza Sabzevari and Jeffrey Delmerico

Eren Allak

Semester Thesis

2015

Eren worked on a novel interest point detector and descriptor from the normal maps generated by Photometric Stereo. Such interest points are used to match texture-less objects in order to register Photometric Stereo reconstructions from different views and removing their intrinsic geometric ambiguities.

Supervised by Reza Sabzevari

Andreas Forster

Semester Thesis

2015

Andreas worked on active camera control: he closed the loop between a visual-SLAM system and the shutter speed controller of a camera. The camera controller he developed automatically selects the best shutter speed and gain such that the SLAM system performs best. Vice-versa, the the visual-SLAM system is informed about changes in the camera system to increase the robustness of direct feature tracking.

Supervised by Christian Forster and Manuel Werlberger.

An-Phi Nguyen

Semester Thesis

2015

An-Phi worked on Dense Multi-Body Structure from Motion. He developed a pipeline to segment images from a perspective monocular camera in order to estimate the camera ego-motion, recover the scene structure and estimate the eoru-motion of the other moving objects together with their 3D structures.

Supervised by Reza Sabzevari

Nathan Baumli

Master Thesis

2015

Nathan worked on evasive maneuvers with quadrotors using dynamic vision sensors (DVS). His algorithm can detect and track an object thrown at a quadrotor and predict whether a collision will occur.

Supervised by Elias Mueggler and Flavio Fontana

Astrid Schlestein

Master Thesis

2014

Astrid designed an automatic system identification routine to determine dynamical models of quadrotors using a motion capture system. The identified dynamical model is then used for delay compensation in our quadrotor control algorithms.

Supervised by Matthias Faessler and Flavio Fontana

Jonas Schuler

Master Thesis

2014

Jonas developed an infrastructure to study the opportunities that photometric 3D reconstruction methods bring to Robotics. He designed a mobile image acquisition system for dense reconstruction of high detailed surfaces.

Supervised by Reza Sabzevari

Maximilian Schulz

Semester Thesis

2014

Maximilian developed a state estimator that enables autonomous, vision-based quadrotors to land safely in case visual tracking fails. His work was the first step in making our quadrotors safe in case of sensor failures.

Supervised by Matthias Faessler and Flavio Fontana

Adrian Rechy Romero

Semester Thesis

2014

Adrian developed a framework to enable in-flight calibration of sensor biases and rotor thrusts for our quadrotors. His framework is now used on a daily basis whenever we fly quadrotors.

Supervised by Flavio Fontana and Matthias Faessler

Raphael Meyer

Semester Thesis

2014

Raphael designed an electrical and mechanical adapter board for quadrotors. It is tailored to the requirements at the Robotics and Perception Group and started the development of an entire custom quadrotor platform.

Supervised by Flavio Fontana and Matthias Faessler

Quim Sanchez

Master Thesis

2014

Quim investiagted fast vision-based relocalization to recover the pose of a micro aerial vehicle after an agile maneuvre, e.g. a flip, in a previously built map. He developed various relocalization schemes for the SVO algorithm that are available open-source.

Supervised by Christian Forster and Matia Pizzoli

Nathan Baumli

Semester Thesis

2014

Nathan worked on lifetime estimation of events from Dynamic Vision Sensors (DVS). A direct application of this is the construction of sharp gradient images at any time instant. His work was presented at ICRA 2015 in Seattle. [ Paper ]

Supervised by Elias Mueggler and Christian Forster

Amos Zweig

Semester Thesis

2014

Amos developed the first event-based 3D reconstruction algorithm assuming known camera trajectory and a reference grayscale image.

Supervised by Elias Mueggler and Christian Forster

Basil Huber

Master Thesis

2014

Basil implemented an event-based tracking algorithm to estimate the pose of a quadrotor during high-speed maneuvers. He also developed a tool to capture ground-truth data from the Dynamic Vision Sensor (DVS). We presented his work at IROS 2014 in Chicago. With his thesis, he won the 2014 Fritz Kutter Award for Industry Related Thesis in Computer Science. [ Paper ] [ Video ]

Supervised by Elias Mueggler and Davide Scaramuzza

Mathias Weyland

Master Thesis

2014

In the first part of his thesis, Mathias developed a custom board to attach a Odroid-U2 computer to the NanoQuad micro aerial vehicle. The board contains a level-shifter circuit, a DC/DC converter and a high-speed USB hub. Secondly, Mathias helped to formalize the photometric disparity uncertainty for next-best-view planning in 2D that we used in the RSS'14 paper.

Supervised by Christian Forster and Matia Pizzoli

Karl Schwabe

Master Thesis

2013

Karl developed a monocular pose estimation system based on infrared LEDs. We released his system as open source and presented it at ICRA 2014 in Hong Kong. [ Paper ] [ Video ] [ Code ]

Supervised by Matthias Faessler, Elias Mueggler, and Davide Scaramuzza

Benjamin Keiser

Master Thesis

2013

Benjamin implemented a torque controller for the KUKA youBot arm to follow end-effector trajectories both smoothly and fast. We released his controller as open source. His thesis won the KUKA Best Student Project Award in 2013! [ Thesis ] [ Video ] [ Code ]

Supervised by Matthias Faessler and Elias Mueggler

Christian Krebs

Master Thesis

2013

Christian developed a system for localizing an aerial-robot in a ground-robot's map using dense full-image alignment. We used a texture-augmented SDF volume as a map representation and showed that the system works with radically different viewpoints.

Supervised by Christian Forster and Matia Pizzoli

Oliver Brand

Master Thesis

2013

Oliver developed a Kinect-aided robust gesture recognition system for human-robot interaction. We used his system at the Robots-on-Tour festival to show interaction with flying robots.

Supervised by Davide Scaramuzza and Max Lungarella

Fabrizio Cortesi

Semester Thesis

2013

Fabrizio developed a method for visual homing in three dimensions that does not require a map or any depth information.

Supervised by Christian Forster and Davide Scaramuzza

Daniel Heid

Master Thesis

2013

Daniel developed two different vision-based 3D reconstruction systems. The first was to reconstruct both medical pillbox and pills and the second was to reconstruct larger objects actively with a camera equipped ground-robot.

Supervised by Davide Scaramuzza and Max Lungarella

Yves Albers-Schoenberg

Semester and Master Thesis

2013-2014

Yves worked with Andras on localizing a flying robot in a city using a single camera onboard the robot and a database of street-view images collected by Google. His system reached remarkable performance despite radical view-point and season differences. His work was published at ICRA 2013 and IROS 2014.

Supervised by Davide Scaramuzza and Andras Majdik

Tim Oberhauser

Semester Thesis

2013

Tim developed a system to estimate the camera pose in a known map previously built by an RGB-D camera attached to a ground-robot. This system is intended to be used for air-ground robot collaboration.

Supervised by Christian Forster and Davide Scaramuzza

Mitchel Bakker

Semester Thesis

2013

Mitchel worked on the air-ground robot collaboration project and implemented a system to estimate the aerial robot pose relative to the ground-robot using AR markers.

Supervised by Christian Forster and Davide Scaramuzza

Maurice Göldi

Semester Thesis

2013

Maurice studied the EKF formulation of the visual SLAM problem and implemented a SLAM simulator in MATLAB.

Supervised by Christian Forster and Davide Scaramuzza

Jonas Strubel

Semester Thesis

2013

Jonas developed a system to track blinking LEDs with the Dynamic Vision Sensor (DVS).

Supervised by Andrea Censi, Davide Scaramuzza, and Christian Brändli

Jan Portmann

Semester Thesis

2013

Jan developed a system to interface with the Optitrack motion capture system using the Robot Operating System (ROS) and Matlab. Additionally, he implemented a system to compare trajectories.

Supervised by Volker Grabe and Christian Forster

Dominik Schlegel

Bachelor Thesis

2013

Dominik developed a tracking device with two servo motors and one RGB camera to estimate the pose of a flying robot relative to the Turtlebot ground-robot.

Supervised by Christian Forster and Davide Scaramuzza

Christian Saner

Semester and Master Thesis

2012-2013

Christian was our first student and built our first micro aerial vehicle based on the AR.Drone. He was a great help at many demonstrations, including the Robots on Tour festival, where he showed his Macarena dance demo of multiple MAVs.

Supervised by Christian Forster and Davide Scaramuzza