Quicklinks

Varcity 3D Challenge online!
The VarCity 3D Semantic Segmentation Challenge based on our ECCV 2014 paper is now online for data and evaluation code!

Varcity Videos!
The VarCity Video Collection based on our previous papers is online!

News

May, 19th 2017: World Premiere!
On that magic day, we will show the world premiere of VarCity - The Video!

April, 25th 2017: Siggraph accepted!!
Our talk on Struggles and Triumphs of The VarCity Video was accepted to Siggraph 2017! :)

April, 1st 2017: Journals online!
Our two CVIU journals for efficient surface meshes and architectural structural element decomposition are now online!

January, 6th 2017: Papers online!
All papers and additional material are updated and online now!! Pfew, 2016!

December, 6th 2016: ICPR Award - BIRBA!
Our Street and Aerial Fusion paper received the ICPR 2016 - Best Industry Related Paper Award (BIRPA) runner-up prize!

October, 13th 2016: Landmarkness code!
The code for the Kobyshev score for 3D saliency for landmarks is now available!

September, 16th 2016: Project Meeting!
Project and collaborators meeting is on at Villa Hatt, Zurich!

September, 2nd 2016: Latex PhD template!
The latex phd template for my PhD thesis is now available!

August, 31st 2016: 3DV oral!
Our paper was accepted as oral at International Conference on 3D Vision (3DV) 2016, Stanford, USA.

July, 13th 2016: ICPR orals!
Two papers were accepted as orals at International Conference on Pattern Recognition (ICPR) 2016, Cancun, Mexico.

June, 10th 2016: CVIU!
Proud to see that our two CVIU - Special Edition for Large-Scale 3D Modeling were both accepted!

May 19th, 2016: In the NEWS!
Michael's CVPR 2016 paper Video2GIF: Automatic Generation of Animated GIFs from Video
made it into the news twice! Once at motherboard and once at leMonde!

April 14th, 2016: Demo Visit!
Demonstration exchange with Margarita Chli, Vision for Robotics Lab, ETH Zurich.

March, 9th, 2016: Papers at CVPR 2016 accepted!
Four papers were accepted at CVPR 2016, Las Vegas, USA.


March, 7-9th, 2016: Papers at WACV 2016!
Four papers were presented at WACV 2016, Lake Placid, USA.


February, 18th 2016: Project Meeting!
Project and collaborators meeting at Holodeck, ETH Zurich!


January, 27th 2016: Project Meeting!
Project and collaborators meeting was held at Stadtgaertnerei, Zurich!


December, 9th 2015: OpenLabNight!
The VarCity projects including the first full city game fly-over are shown during the OpenLabNight @ Computer Vision Lab, ETH Zurich!


Nov, 8th 2015: 3D Cities for Autonomous Cars?
VarCity was discussed during the Dagstuhl Seminar for Vision for Autonomous Vehicles and Probes at Schloss Dagstuhl, Germany.

Sep, 21th 2015: Showcase!
VarCity was showcased by the ERC Funding Agency.

Sep, 14th 2015: Presentation!
VarCity was presented and discussed at INRIA research group at INRIA Sophia Antipolis Mediterranee, France.

June, 30th 2015: Presentation!
VarCity was presented and discussed during a PhD examination at IMAGINE research group at the Universite Paris Est, Paris, France.

June, 11th 2015: Workshop Meeting!
VarCity was presented at the Workshop for Semantics for Visual Reconstruction and SLAM at CVPR 2015, Boston.

May, 11th 2015: Project Meeting!
Project and collaborators meeting was held jointly with the CVL Workshop at Schloss Rapperswil, Switzerland and special guest star
Richard Hartley!

April, 29th 2015: Presentation!
VarCity was presented at archilogic.

March, 19th 2015: Presentation!
VarCity was presented at University of Bonn, Germany.

March, 4th 2015: Presentation!
VarCity was presented at Toshiba Corporate Research & Development Center, Toyko, Japan.

March, 2nd 2015: Papers at CVPR 2015 accepted!
Five papers were accepted at Computer Vision and Pattern Recognition (CVPR) 2015, Boston.

December, 3rd 2014: Demo Visit!
Demonstration exchange with
Davide Scaramuzza, Robotics and Perception Group, University Zurich.

December, 2nd 2014: Project Meeting!
Project and collaborators meeting was held at Villa Hatt, ETH Zurich.

November, 23rd 2014: Workshop!
Participating in the ETH-MPI Computer Vision Workshop by the Max Plack Institut for Intelligent Systems, Tuebingen, Germany.

October, 1st 2014: 3D Challenge online!
The VarCity 3D Semantic Segmentation Challenge based on our ECCV 2014 paper is now online for data and evaluation code!

September, 12th 2014: PhD Defense!
Dr. Julien Weisseberg has finished his PhD!

September, 8th 2014: Papers at 3DV 2014 accepted!
Three papers were accepted at 3DV 2014, Tokyo.

September, 5th 2014: Conference in town!
The European Conference on Computer Vision (ECCV) 2014 will be held in our town this year!

August, 28th 2014: Project Meeting!
Project and collaborators meeting was held at Technopark with talks from ESRI / CityEngine, faceshift and smarterbettercities!

July, 8th 2014: Papers at BMVC 2014 accepted!
Two papers were accepted at BMVC 2014, Nottingham.

June, 15th 2014: Papers at ECCV 2014 accepted!
Four papers were accepted at ECCV 2014, Zurich.

April, 3rd 2014: Project Meeting!
Project and collaborators meeting is happening in a think tank at ETH Zurich!

February, 24th 2014: Papers at CVPR 2014 accepted!
Three papers were accepted at Computer Vision and Pattern Recognition (CVPR) 2014, Columbus, Ohio.

January, 31th 2014: PhD Defense!
Dr. Alex Mansfield has finished his PhD!

September, 3rd 2013: Project Meeting!
Project and collaborators meeting is on at St. Peter Church, Zurich!

August, 29th 2013: Papers at ICCV 2013 accepted!
Six papers were accepted at International Conference on Computer Vision (ICCV) 2013", Sydney, Australia.

August, 15th 2013: Architecture meets Vision meets Architecture Meeting!
Interactions between architecture and computer visionn workshop was held at CVL Seminar, ETH Zurich!

July, 1st 2013: Paper at BMVC 2013 accepted!
Our work on view clustering for reconstructions was accepted at British Machine Vision Conference (BMVC) 2013, Bristol, UK.

May, 15th 2013: Swiss Vision Day!
The VarCity project and its publications are shown at the Swiss Vision Day @ ETH Zurich! Lots of posters, demos, food, and fun. Join us again next year!

May, 08th 2013: Project Meeting!
Project and collaborators meeting is on today at ETH Zurich!

February, 25th 2013: CVPR!
We have three accepted papers for CVPR this year. More details later...

January, 18th 2013: Project meeting!
Project and collaborators meeting is on today at Villa Hatt, ETH Zurich!

December, 19th 2012: OpenLabDay!
The VarCity project are shown at the
OpenLabDay @ Computer Vision Lab, ETH Zurich! Lots of demos, food, and fun. Join us again next time!

November, 12th 2012: SPAR 2012 Demo!
The VarCity project demos are shown at
SPAR 2012 conference, The Hague. Prof. Luc Van Gool presented the invited keynote speech.

October, 7th 2012: ECCV and 3D PVT 2012!
Six papers presented at
ECCV 2012 and 3DPVT 2012.

September, 14th 2012: Project meeting!
Project and collaborators meeting went on today at HIT ValueLab, Hoenggerberg! The ValueLab is an environment for collaborative design and management.

September, 1st 2012: Website online!
First version of the VarCity project website is online! More (re)construction soon!

August, 15th 2012: Kickoff meeting!
Kickoff meeting for all partners and people involved went on today at the Zurich Zoo!

May, 1st 2012: Launch at ETH Zurich!
Today was the launch day at ETH Zurich in the Computer Vision Lab.

This page has been edited by Hayko Riemenschneider.

Locations of visitors to this page Locations of visitors to this page

VarCity - semantic and dynamic city modelling from images

Computer Vision Laboratory, ETH Zurich

Semantic cities modelled on knowledge are combined with dynamic cities containing events and traffic flows!
VarCity - The Video - full video available May 22nd 2017 here!



VarCity was a multi-year research project financed by the European Research Council and obtained by ETH Professor Luc Van Gool 5 years ago at the Computer Vision Lab, ETH Zurich. Dr. Hayko Riemenschneider was appointed to lead the group of researchers on a daily basis, see people for all names.

VarCity - The Video showcases some of the buildings blocks of creating an entire city from images.

For press information, questions, usage of the video, collaborations or any further contact, you can reach us at
hayko(at)vision.ee.ethz.ch [EN,DE] or vanhoey(at)vision.ee.ethz.ch [EN,FR]

You can also follow the new account on twitter: https://twitter.com/VarCityETH










Current results in 2017!
Semantic cities modelled on knowledge are combined with dynamic cities containing events and traffic flows!
(c) VarCity - Semantic Cities (c) VarCity - Dynamic Cities


3D city models have many applications, such as urban design, navigation, real-estate ads or movies & video games. Our project aims at producing such models from photographs both faster and with higher detail than before. As the effort often needs to be repeated to regularly update a city model, efficient production of compact models is an absolute necessity.
We process images of real cities automatically and efficiently to create parameterized and semantic 3D models, in which streets, buildings and vegetation are discriminated, and the number of floors, positions and shapes of windows, doors and balconies are recognized and encoded.
Our research further fills our static 3D city models with dynamic content by extracting special events and traffic flows from images, and by generating a city-scale motion and activity model. One can virtually visit the Munsterhof in Zurich and see a video summary of recent events there, or check out traffic densities along the kids' way to school.










Vision in 2012!
Semantic cities modelled on knowledge are combined with dynamic cities containing events and traffic flows!
(c) Procedural, CityEngine. (c) Trey Ratclif, CityScape.


Virtual city models are used in many game and movie designs, like the industry leading spin-off Procedural (now part of ESRI) for creating stunning 3D urban environments from 2D data. Currently the production of real 3D city models comes at a high cost. Given that the modeling effort needs to be repeated regularly for updating, rendering city model production more efficient is an absolute necessity. Our work creates inverse procedural models, which are built for existing cities. The modeling is done by analyzing images of real cities and constructing parametrized and semantic models, where we know the number of storeys, shadows cast by new buildings, the position of traffic signs, vegetation, etc.
Our research further creates dynamic living 3D city models, which allows for deeper immersion than in current city representations. We extract special events and traffic flows to generate a city-scale motion and activity model. One can virtually visit Times Square and see what was on the electronic newsreel recently or check out traffic densities along a journey or the kids' way to school.