From September 18, 2022 to September 22, 2022, INTERSPEECH 2022, an international conference dedicated to AI voice and signal processing was held in Songdo, Incheon.
INTERSPEECH is the world’s largest voice processing technology conference held by the international Speech Communication Association (ISCA). This year’s conference marks its 23rd anniversary! Speech signal processing, speech recognition, speech synthesis, multilingual translation, speech conversion, and assistive speech therapy are required for human-to-human, human-to-object and object-to-object language and voice communication were discussed this year under the theme of “Human and Humanizing Spoken Language”.
Speech signal processing, speech recognition, speech synthesis, multilingual translation, speech conversion, and assistive speech therapy are required for human-to-human, human-to-object and object-to-object language and voice communication were discussed this year under the theme of “Human and Humanizing Spoken Language”. 
(LG’s booth at INTERSPEECH 2022!)
(A lot of those who attended the conference stopped by our booth!)
LG AI Research, LG Electronics and LG Uplus did a joint LG Booth and introduced out voice recognition technology.
LG AI Research demonstrated the YouTube news video STT (speech-to-text) function at the LG integrated booth. The STT function is a technology that converts the voice from YouTube News into text in real time and extracts keywords from the text. Applying this technology to customer service can increase the productivity by helping customer service agents quickly retrieve corresponding service manuals or refer customers to different departments instantly.
The LG team also introduced a technology that can summarize the key issues of the customer inquiry. The team demonstrated how the technology can classify customer inquiries into different service categories such as complaint or refund requests, and the customer’s emotional state into positive, negative and neutral.
(Our YouTube News demo at the booth!)
(Our Language Lab Leader, Stanley Choi gave an Industry Talk at INTERSPEECH 2022)
)Stanley Choi, Leader of the Language Lab presented and discussed EXAONE’s advancements in speech and language technology during the conference’s Industry Talk Event. Self-supervised learning, E2E (End-to-End) voice technology, optimization and domain application are the four phases of EXAONE voice reseach. Stanley stressed the significance of the individual research phases, stating that the research into optimization enables the technology to be implemented with greater precision, faster inference and lower computing costs. He also shared positive research outcomes that LG AI Research has made, explaning that LG AI Research developed GPU card-less real-time streaming E2E STT deduction technology, and that an internal evaluation revealed that EXAONE demonstrates a stable voice recognition accuracy. It was a good opportunity for us to find out about the latest research conducted by the Language Lab and how the research outcomes can impact AI research in the future.
LG AI Research plans to participate in more conferences int he latter half of this year! We look forward to your interest and support!