Government will add 190 types of AI learning data to AI Hub in six major areas, including self-driving and health care

The Ministry of Science and ICT announced that it will add and open 190 types of artificial intelligence learning data (580 million cases) that will drive the development of artificial intelligence technology and industries in Korea through the AI hub from July 12.

The Ministry of Science and ICT has been building and offering AI learning data since 2017, which is difficult for companies and researchers to secure individually due to time and cost issues and has been recording more than 180,000 downloads after significantly increasing the construction size since 2020.

The data will be sequentially opened after verifying the quality and utilization of 190 types of data established in 2021, in addition to 170 types of data (480 million cases) in 2020, which was opened in June last year, and the total size will reach 381.

The 190 types of AI learning data in six fields to be opened this time are the result of the participation of industry, experts and the public from planning to construction.

Based on private demand, industry, academia, and research experts and data utilization companies directly participate in the project to discover and select data that has a large industrial ripple effect and is difficult to build on a large scale in the private sector.

The date includes △ Korean voice and natural language (44 kinds of children, multi-talker voice), △ video and image (39 kinds of sports movements, pets, etc.), △ traffic and logistics (22 kinds of robot driving, traffic accidents, etc.), △ healthcare (32 kinds of heart disease, and mental health), △safety & environment (27 kinds of overloaded vehicles, logistics, etc.), △ agro-livestock-fisheries (26), etc.

In particular, in the data construction process, a total of 811 companies and institutions including major domestic artificial intelligence and data companies, universities (118 including Seoul National University, KAIST, etc.), and hospitals (50 including Seoul National University Hospital and National Cancer Center) and up to 43,000 people participated.

In addition, in order to secure the level of quality required by actual data consumers, along with professional quality verification by the Telecommunications Technology Association(TTA), which standardizes and certifies information and communication technologies, it has reviewed and supplemented data utilization by large companies such as NAVER, Kakao, LG, KT, and Hyundai Motor Company, startups such as RideFlux and StradVision, universities and research institutes such as GIST, Pusan National University, ETRI, and Korea Automotive Technology Institute.

The Ministry of Science and ICT and the National Information Society Agency plan to continuously manage and improve data quality and utilization through public-private cooperation, such as operating a regular window for collecting opinions on quality in AI ​​Hub, and supporting data quality improvement research in connection with the Korean Institute of Information Scientists and Engineers (July~December), even after data is opened.

Meanwhile, with the opening of the data, industry, academia, and related experts held ‘the 3rd Artificial Intelligence Data Utilization Council’ under the 2nd Vice Minister of Science and ICT Park Yoon-kyu to discuss ways to promote AI Hub data utilization and spread performance, and collected vivid opinions from artificial intelligence, data industry, and research sites.

Reported by Smart City Today