메인콘텐츠 바로가기

ETRI Webzine

VOL.51
April 2019

ETRI News

News  ____  ETRI Issue

페이스북 공유하기페이스북 공유하기 트위터 공유하기트위터 공유하기 카카오톡 공유하기카카오톡 공유하기

Headline News

ETRI Distributes Five-language Voice Database Containing English-Language Sentences

ETRI researchers established a high-quality voice database in five languages, including Thai, Malay, Indonesian, Arabic, and Vietnamese languages, which also contains the corresponding sentences translated into English. This will help reduce the dependence of domestic companies on foreign technology in various voice recognition and translation services applied to AI speakers, navigation systems, and IoT solutions. Also, this achievement will contribute to the further development of various services using those languages.

ETRI Develops Capsule Endoscope Based on Human Body Communication Technology

Using human body communication technology, ETRI researchers, jointly with a domestic company, developed a capsule endoscope capable of effectively diagnosing the esophagus and stomach, which account for nearly 54% of digestive diseases.
ETRI announced that the application of in-body communication technology quadrupled the existing image transmission rate implemented by a domestic company, i.e., enabling a high-speed transmission at a rate of 24 frames per second.

ETRI Serves as Priming Water for AI Services

Through its innovative growth engine project targeting the AI field, the Ministry of Science and ICT is pushing forward its initiative to induce the disclosure of the AI technologies developed by ETRI researchers and thus enable anyone to develop AI-related services. By doing so, the researchers are planning to promote the application and development of various AI solutions.
Key information disclosed by ETRI includes language, voice, and visual intelligence APIs and data. This helps domestic SMEs, researchers, and developers to develop AI-based services more efficiently. ETRI researchers have thus far disclosed data related to Q&A, dialogue processing, object recognition, and voice recognition (multilingual).