Soniox Launches New Speech AI Models
Soniox has released speech recognition artificial intelligence technology with new foundational AI models, the Soniox mobile app, and Soniox Playground, allowing users to experience the new era of voice AI firsthand.
Soniox's models will originally be available for nine languages, starting with English and Korean.
In developing the technology, Soniox built a ground-up infrastructure to process and train large models on massive amounts of audio and text. Specifically, Soniox processed more than 1 million hours of audio data for training. The entire training process was completed on a single A100 server in less than four weeks.
Soniox also had to design new and more efficient neural network architectures and develop new criterions that inherently prioritize low-latency decision-making while still considering accuracy.
"Achieving human-parity or superhuman accuracy is of paramount importance, rather than settling for a solution with a misrecognition rate of 20 percent or higher, which proves to be useless for most applications," said Klemen Simonic, founder and CEO of Soniox, in a statement.