Netmarble will introduce deep learning-based mobile voice recognition technology at 'NeurIPS 2020 (Neural Information Processing Systems)', which will be held in Vancouver, Canada from December 6.
NeurIPS, the world's most prestigious AI conference, deals with research in various AI fields, including machine learning, big data and audio-visual information processing.
At the conference demonstration session, Netmarble will unveil "MONICA: MObile Neural voIce Command Assistant for mobile games" to the extent that deep learning-based voice recognition devices are available in mobile games for the first time in the world. "MONICA" passed the screening with perfect scores after receiving a review result of "Strong Accept" from all the paper examiners.
"We are focusing on continuously improving the voice user experience (UX) that is needed for mobile games because recognition rate and response speed are very important," said Ahn Soo-nam, head of the AI Center at Netmarble. "We are focusing on improving the immersion of game users by developing the world's first voice recognizer that responds with high accuracy and fast response speed within mobile devices."
"We expect to provide better convenience to game users through the development of voice recognition technology operated by mobile devices," said Oh In-soo, head of the AI Center at Netmarble. "We plan to apply the technology sequentially to global titles that will be released by Netmarble in the future."
Meanwhile, Netmarble is currently providing starter recognition technology (starting language: Monica) to its 'A3: Still Alive' released in March.
As a paper at INTERSPEECH 2019, the world's top academic society in the AI field, and as a paper at "IEEE ICASSP 2020," the technology presentation was conducted to prove Netmarble's technological prowess in the artificial intelligence field.
