ANDREA CENSI PHD THESIS

If you are interested in doing a custom student research project, please email the project supervisor of your choice directly. February 28, RPG featured in the Weltwoche magazine Weltwoche, a Swiss weekly magazine, talks about drones in the daily life. For more details, see the ICRA’15 paper and the accompanying video. Welcome to Henri Rebecq as a new PhD student in our lab! We believe that event cameras will allow future robots to move faster and more agilely.

Titus Cieslewski, a PhD student in our lab, has received a prize from Homegate during the HackZurich hackathon, for the project Wonsch. Custom Projects From time to time, project supervisors will develop custom student research projects to fit with a student’s particular interests or skills. Read the paper for further details. Our research on autonomous drone racing was featured on NewScientist. More details in our SSRR’14 paper.

Robot Perception Group

Welcome to Davide Falanga as a new PhD student in our lab! For more details, see the ICRA’15 paper and the accompanying video. Other portraits by or andreea robots and roboticists.

The lifetime of an event is the time that it takes for the moving brightness gradient causing the event to travel a distance of 1 pixel. This is the first ever autonomous flight with an event camera, which demonstrates agile maneuvers and flying in low-light environments.

andrea censi phd thesis

Then he ended up programming robotic cars. Our work on the foldable drone, the first quadrotor able to change morphology in flight to adapt its shape and size to different tasks, won the Drone Hero Award Contest for the category Innovative Drone.

  ANG KAHALAGAHAN NG PAGPAPLANO NG PAMILYA ESSAY

andrea censi phd thesis

San Francisco, CA, September We also provide an interface to the RotorS Gazebo plugins to use our algorithms in simulation. November 1, Four new members We welcome Dr.

Here is a documentary piece ansrea Red Hat, “The road to AI”, that happens to talk about two of the things that keep me busy: Please use our form to send your comment to us: Exploiting motion priors in visual odometry for vehicle-mounted cameras with non-holonomic constraints. For more details, check our research page.

We hope the list will help newcomers to the field to get started with this technology by directing them to the appropriate references. Check out the video here. June 1, New Postdoc We welcome Dr. Have a look at the gallery!

The goal of bootstrapping is creating agents that are able to learn “everything” from scratch, including a torque-to-pixels models for its robotic body. Details are in our ISER’16 paper. July 14, Binaries for SVO 2. RPG is very proud of them!

Bootstrapping Vehicles: A Formal Approach to Unsupervised Sensorimotor Learning Based on Invariance

This paper shows that, for the case of localization with range-finders, Fisher’s matrix is a function of the expected readings and of the orientation of the environment’s anfrea at the sensed points.

  LANCIA THESIS RECENZE OJETIN

Our paper about safe quadrotor navigation computing forward reachable sets was accepted for publication in the Robotics and Automation Letters RA-L The 2 km dataset consists of time synchronized aerial high-resolution images, GPS and IMU sensor data, ground-level street view images, and ground truth data.

Check out our research page on event-based vision. Homepage Navigation Content Sitemap Search.

June 13, Paper accepted in RA-L ! Davide Scaramuzza talks about autonomous, agile, vision-controlled drones and event cameras at the GeorgiaTech robotics seminar series: We combined deep networks, local VIO, Kalman filtering, and optimal control. An ICP variant using a point-to-line metric.

Check out the media coverage here. We are interested in your comments.

andrea censi phd thesis

Welcome to Alessandro Simovic as a new drone engineer in our lab! Our recent work on decentralized visual place recognition using a distributed inverted index was accepted to RA-L! This is the first instance of a bootstrapping agent that can learn the dynamics of a relatively large universe of systems, and use the models to solve well-defined tasks, with no parameter tuning or hand-designed features.

ANDREA CENSI PHD THESIS

RPG is very proud of them! September 16, New visiting postdoc Welcome to Dr. This paper concerns the introduction of uncertainty in the MCDP framework. Check out the article here. A demonstration of the capabilities of his controller is shown in this video.

For more information, read the ETH information on plagiarism. We combined deep networks, local VIO, Kalman filtering, and optimal control. To obtain more agile robots, we need faster sensors. Homepage Navigation Content Sitemap Search. Our approach fuses deep learning and optimal control to achieve the ultimate flight performance. If you are interested in doing a custom student research project, please email the project supervisor of your choice directly.

PaperYouTubeCode on Github. April Last update of research info: Citation Censi, Andrea Bootstrapping Vehicles: Half of the talent in the world belongs to women, but unfortunately they are still the minority in robotics. Tgesis thanks to Google!

Theses & Semester Projects – Institute for Dynamic Systems and Control | ETH Zurich

If you are interested in doing a custom student research project, please email the project supervisor of your choice directly. How do you fulfil your career aspirations? ESIM readily provides ground truth depth and optic flow maps.

In particular, I am responsible for creating the conference trailerwhich involves distributing thousands of rubber duckies to all roboticists in the world.

andrea censi phd thesis

Event cameras allow predicting the steering angle of a car more robustly and accurately at night and high dynamic range scenes than a standard camera. Motion planning in observations space with learned diffeomorphism models.

  HIUS 221 READING QUIZ 1 ESSAY

It can squeeze to fly through narrow gaps. Check out the video here nadrea, and the paper here.

For more information, please see the site co-design. Our recent work on real-time parallel tracking and mapping with an event camera was accepted to RA-L!

andrea censi phd thesis

Learning to Fly by Driving We have deviced a Deep Neural Network, called DroNet, that teaches a drone how to fly autonomously and safely in the streets of a city, among other vehicles, by imitating the behavior or cars and bicycles! Our interests encompass both ground and micro flying robots, as well as multi-robot heterogeneous systems consisting of the combination of these two.

Bootstrapping Vehicles: A Formal Approach to Unsupervised Sensorimotor Learning Based on Invariance

The video shows learning of a bilinear model of a sensorimotor cascade for a camera. The Duckietown project, developed at ETH Zurich, allows everybody to pbd how self-driving cars work.

andrea censi phd thesis

In this scenario, node expansion is very costly, as each node in the graph is associated to an uncertain diffeomorphism and corresponding predicted observations. PAMPC combines control and planning in one solution and allows to not only achieve an action objective but also compromise it with a perception objective. We release the first public, large-scale dataset recorded with a drone in anrdea urban environment at low altitudes m.

  ANG KAHALAGAHAN NG PAGPAPLANO NG PAMILYA ESSAY

This is the first instance of a bootstrapping agent that can learn the dynamics of a relatively large universe of systems, and use the models to solve well-defined tasks, with no parameter tuning or hand-designed features. Philipp Foehn, PhD student in our lab, and Naveen Anrdea, former visiting researcher, received the Best Student Paper Award Finalist prize at RSS in Boston for our work on trajectory optimization for agile quadrotor maneuvers with cable-suspended payloads.

Check out our media andrae In this clip, we summarize our main achievements, projects, awards, exhibitions, and upcoming videos!

Andrea censi phd thesis

For more information, read the ETH information on plagiarism. Andea you be able to learn a model of your body and use it to perform useful tasks? Some highlights are shown in a video by Le Matin. More information and software credits. More details in our SSRR’14 paper. April 12, Davide Scaramuzza gives an invited seminar at Princeton University. It took me ten years to delegate this issue.

The code provided is implemented in Python and produces paper-ready plots and videos for event-based feature tracks. More details are available here.