Upcoming ICCV Workshop: Artisense & TUM to organize 2nd edition of ‘Map-based Localization for Autonomous Driving’ Workshop in Oct, 2021
Artisense and the Technical University of Munich (TUM) will be organizing the 2nd edition of the Workshop on ‘Map-based Localization for Autonomous Driving’ at the International Conference on Computer Vision (ICCV), taking place 11-17 October 2021.
Success at ECCV 2020
This comes as a result of the successful first workshop on ‘Map-based Localization for Autonomous Driving’ (MLAD), which took place at the European Conference on Computer Vision (ECCV) last August 2020. The workshop was organized by the researchers from Artisense and TUM (Daniel Cremers, Rui Wang, Patrick Wenzel, Nan Yang & Niclas Zeller).
The workshop focused on discussions centered on the importance of map information for the problem of autonomous driving, as well as the concomitant challenges of accurate localization within those maps. While most current solutions rely solely on the integration of GNSS and inertial measurements, more sophisticated techniques are required to be able to localize in any environment, including tunnels, urban canyons, etc.
To discuss these challenges, as well as the current state of the art in autonomous driving and mobile robotics, experts in academia and industry shared their personal insights and experiences at the event. This included: Dengxin Dai (ETH Zurich), Andreas Geiger (University of Tübingen, Max Planck Institute), Stefan Leutenegger (Imperial College London, now Technical University of Munich), Cyrill Stachniss (University of Bonn), Raquel Urtasun (University of Toronto, Uber ATG, now Waabi).
In addition to the invited talks, Artisense presented their current research activities in the field of re-localization and unveiled the 4Seasons Dataset, a new multi-weather, all-seasons dataset recorded using the Artisense’s Visual Inertial Navigation System (VINS). The aim of this dataset is to enable research in robust vision-based odometry, as well as map-based localization.
Based on the 4Seasons dataset, the workshop also hosted a re-localization challenge, which was sponsored by Kudan Inc. and Artisense. Paul-Edouard Sarlin from the public research university, ETH Zurich was selected as the winner of the challenge.
The Workshop will include a re-localization challenge based on the 4Season dataset, created by TUM & Artisense.
Upcoming Workshop at ICCV 2021
Over the last year, the 4Seasons dataset continues to garner the interest of those in academia. It has established itself as part of a valuable collection of data for the development and evaluation of algorithms ranging from robust visual odometry under challenging conditions, to place recognition and re-localization across different seasons and weather conditions.
Despite the progress over the last few years, there still remain numerous questions in the field of map-based localization. This includes the ability to efficiently and at a low cost generate maps at a very large scale and more importantly, how those maps can be kept up-to-date.
To explore and answer these questions, Artisense and TUM will be organizing the second edition of the MLAD workshop at this year’s virtually held ICCV event, happening on 11 October 2021.
Upcoming workshop and challenge at ICCV 2021, organized by TUM, Artisense & Kudan Inc.
Confirmed speakers for this workshop include Wolfram Burgard (University of Freiburg, Toyota Research Institute), Michael Milford (Queensland University of Technology) and Torstens Sattler (Czech Technical University), with several more speakers to be expected.
Furthermore, the workshop will once again host the re-localization challenge based on the 4Seasons dataset that will feature more challenging real world scenarios, ranging from urban to country side environments. The event is sponsored by Kudan Inc. and Artisense.
For more details on the workshop and topics covered, visit here.
About Artisense
Artisense is a computer vision and sensor fusion software company that develops an integrated localization platform using cameras as lead sensor for the automation of robots, vehicles and spatial intelligence applications. Artisense provides products and technology for highly accurate, robust and safe navigation or surveying in any space and with low-cost hardware.