Sharing of Naver Labs' Proprietary High-Precision Map Data
[Asia Economy Reporter Jin-gyu Lee] Naver Labs announced on the 9th that it will hold the 'Naver Labs Mapping & Localization Challenge' to share mapping and localization technology and data with domestic researchers.
Naver Labs is researching technology aimed at future urban environments (A-CITY) where movement within urban spaces is automated using autonomous machines. As part of this, it is advancing high-definition maps (HD maps) and high-precision localization technology. A representative example is the R2D2 technology, which ranked first at 'CVPR 2019,' the world's most prestigious conference in computer vision and deep learning, surpassing global IT companies.
This challenge was organized to revitalize research in visual localization technology and related fields by sharing high-precision map data independently built by Naver Labs with domestic technologists. Participants select one or more indoor/outdoor datasets containing arbitrary photos and mapping space data provided by Naver Labs, then submit the specific location information values where the photos were taken. There are no restrictions on the methodology used to estimate the location information values, and participants only need to update the estimated location information values by June 27.
A total prize of 20 million KRW will be awarded to the six teams that submit the most accurate location information values, and all members of the first-place team will be given the opportunity to participate in the Naver Labs internship program. Baek Jong-yoon, head of the Autonomous Driving Group at Naver Labs, said, "We plan to support university researchers so they can fully focus on their research during the challenge period," adding, "We hope this will be an opportunity to further advance the technological level in this field domestically despite difficult circumstances."
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.
