Vision21Tech
Provided a child-domain speech recognition model for ages 4–7 that accurately transcribes toddler speech in daycare play and daily-conversation environments
Soundustry, a leader in voice-based AI solutions, was founded by ETRI (Electronics and Telecommunications Research Institute)
and delivers customized solutions powered by deep R&D expertise.
We provide Korea's most accurate English speech recognition for Korean speakers,
along with Korea's first free-speech tutoring solution — all tailored to each customer's needs.
Launched the CanvasTalk app
Launched the AsuTalk app beta service
Graduated with Excellence from the Youth Startup Program
(Chairman's Award, KOSME)
Signed an MoU with the Institute of Language and Information Studies, Yonsei University
ETRI-spun-off company; founded as a youth startup
Selected for Korean I-Corps; Grand Prize winner in a startup competition
Seed funding from Bluepoint Partners; selected as Dream Venture Star with DAEKYO
Selected for the TIPS program
Signed an MoU with the Corpus Research Institute, Incheon National University
Natural Language Understanding
Tuned for Korean English-speaking patterns, our system effectively understands free-form natural speech (free talking)
and evaluates it as accurately as a native speaker.
Expert Fluency Assessment, Correction, and Feedback
English education experts from Seoul National University and Yonsei University provide
fluency scores and personalized feedback to boost learners' skills.
Image-based English speaking, correction, content summarization, assessment, and feedback solution for elementary English diagnosis and learning
English vocalization recognition, pronunciation fluency assessment, and feedback solution for adult Korean learners' English sound tuning
English pronunciation recognition, assessment, correction, and feedback solution for elementary English classes and assignments
Domain-Customized Data
We train extensively on both Korean English speech data and non-native Korean speech data.
New-Word Data
We understand free-form natural speech tuned to Korean speakers and assess it as accurately as a native speaker,
and through adaptive learning that incorporates neologisms and proper nouns,
we help learners keep pace with the latest language trends.
Provided a child-domain speech recognition model for ages 4–7 that accurately transcribes toddler speech in daycare play and daily-conversation environments
Provided an elementary-domain speech recognition model for English speaking diagnosis, learning, and classroom use
Provided a speech recognition model tailored to adult Korean learners of English
218 Gajeong-ro, Yuseong-gu, Daejeon 34129, Republic of Korea
#602, ETRI Convergence Technology Research & Production Center
042-384-1043
soundustry.inc@gmail.com
0504-280-1043