[Asia Economy Reporter Seulgina Jo] Cho Kyung-sik, Vice Minister of the Ministry of Science and ICT, observed a demonstration of the 'AI CCTV Safety System' built using AI training data opened through the Data Dam on the 3rd and stated, "This is a good example of AI technology spreading into the daily lives of the public."
According to the Ministry of Science and ICT, Vice Minister Cho attended the 2nd meeting of the ‘AI Data Utilization Council’ held at Daejeon City Hall on the same day and said, "For the success of digital transformation accelerated by the normalization of non-face-to-face interactions, the establishment and utilization of high-quality data in various fields must be supported."
The council meeting was a gathering where industry, academia, and research stakeholders in the AI field explored ways to activate the use of AI training data. Representatives from 12 companies and institutions, including major corporations, SMEs, startups, and related organizations directly developing and utilizing AI services such as LG, Kakao, Hyundai Motor Company, Deepnoid, Viva&ES, Electronics and Telecommunications Research Institute (ETRI), and Gwangju Institute of Science and Technology (GIST), attended. They shared major achievements and utilization cases since the additional opening of data and collected opinions on various dissemination measures to further promote data utilization.
The AI CCTV Safety System mentioned by Vice Minister Cho is a service that detects abnormal behaviors such as passengers falling on CCTV at stations or disabled passengers in wheelchairs and promptly notifies for quick action. It was built and developed by Daejeon Metropolitan Express Transit Corporation using AI training data opened through the Data Dam.
Currently, the government is promoting the construction and opening of over 1,300 types of AI training data by 2025 as a core project of the Data Dam. The Ministry of Science and ICT and the Korea Intelligent Information Society Agency confirmed increased interest in the use of AI training data through performance analysis over about three months after the additional data opening in June, showing a 2.3-fold increase in the average daily number of visitors to the AI Hub and a 2.7-fold increase in downloads compared to before.
Applications are also increasing in public and industrial fields. On this day, the Ministry of Science and ICT shared specific achievements such as 25 cases of companies and institutions participating in data construction achieving sales through the sale of data authoring tools and AI solutions they developed, 61 cases of universities and research institutions publishing research results or papers in world-class conferences and journals, 26 patent applications, and 6 awards in domestic and international competitions.
Additionally, major cases introduced included the autonomous driving pilot service developed by the startup ‘RideFlux (CEO Park Jung-hee)’, which participates in building data for autonomous vehicles, and the healthcare service by ‘MediaZen (CEO Ko Hoon)’, which is developing a voice recognition service for non-face-to-face medical treatment.
The Ministry of Science and ICT plans to continue implementing mid- to long-term tasks such as additional data expansion following immediate actions like labeling supplementation and reclassification for 59 requests for data supplementation, improvement, and expansion received during the concentrated data improvement period until last September.
Experts attending the council shared the recognition that "to ensure proper AI training and improve accuracy, it is most important to create an environment where sufficiently segmented and reliable high-quality data can be secured and utilized by field." They also emphasized the need for expanded government support for data construction difficult to secure in the private sector, continuous improvement of data quality, and support for enhancing usability.
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.
