ETRI has been releasing AI technology since 2017
Release of Language, Speech, and Visual Intelligence APIs and Data, etc.
| Used over 11.64 million times since release

The Electronics and Telecommunications Research Institute (ETRI) announced on the 20th that it is making its self-developed AI technology available so that anyone can develop AI-related services.

Since October 2017, ETRI has been consistently releasing relevant data to lower entry barriers for domestic AI service development and enhance the competitiveness of the AI-related industry.
Key information disclosed by ETRI includes language, speech, and visual intelligence APIs and data. Through this, it supports domestic small and medium-sized enterprises, researchers, and developers in developing AI-based services more efficiently. The research team has previously released data related to question answering, dialogue processing, object recognition, and multilingual speech recognition.

The technologies additionally unveiled this time include Wikipedia question answering, dialogue processing, image object recognition, and multilingual speech recognition. This can be utilized to develop various AI application services, such as AI speakers/assistants, conversational information and consultation services using chatbots, CCTV crime tracking and prevention services, and automatic interpretation services.

This allows the general public, who are unfamiliar with how services are developed using AI, to understand the overall principles of the technology, while enabling developers to simply use the functions and algorithms created by ETRI researchers without the need for programming.
Since November 2017, ETRI has released the Korean language analysis open API for its language intelligence 'Exobrain' and provided machine learning data. In addition, in December 2018, it additionally released open APIs and data for language, speech, and visual intelligence.

ETRI's AI SW API service has been utilized more than 11.64 million times since its release. Usage was most frequent in the order of industry (42%), universities (33%), individuals (19%), and others (6%), and the average daily usage of Open APIs alone reached 23,000.

In the second half of this year, it is expected that the release of additional APIs in areas such as software technologies—including question-and-answer and human attribute detection technologies in the legal field and Vietnamese speech recognition technology—as well as object detection training data, will be helpful to developers of AI-based services related to law.

AI Open APIs and AI training data are provided through the Public AI Open API & Data Service Portal, and any citizen can obtain an account and use them for free for research purposes within the daily allowable usage limit.
The research team anticipates that this API can be used when developing AI products or services to enable voice personal assistant services, automatic interpretation services, intelligent CCTV, and conversational information consultation services.

Hwang Chun-sik, Head of the SW-SoC Convergence R&BD Center at ETRI, stated, “I hope that various AI services will emerge by utilizing the released APIs,” adding, “We will continue to strive to release high-quality AI SW APIs and data.”

This project has been underway since 2017 as part of the 'AI Innovation Growth Engine Project' and the 'Intelligent Information Industry Infrastructure Development' project supported by the Ministry of Science and ICT. style="text-align: center;">
Plan to Expand Public AI Open API and Data Services