Silero STT - Realtime Speech-to-Text

Perform realtime speech-to-text on device with the Silero speech-to-text machine learning model. This package requires NatML to run the machine learning model. Features include: + Bare Metal Performance. The Speech-to-Text predictor is implemented with NatML, which takes advantage of hardware machine learning accelerators, like CoreML on iOS and macOS, NNAPI on Android, and DirectML on Windows. View the model performance statistics on NatML Hub. + Extremely Easy to Use. The Speech-to-Text predictor accepts audio data (AudioClip, NatDevice, or any PCM audio data) and returns the detected text as a plain string. + Cross Platform. The Speech-to-Text predictor supports Android, iOS, macOS, and Windows alike. As such you can test it in the Editor, and deploy it to the device all in one seamless workflow. + Lightweight Package. This package contains the predictor scripts, whereas the ML model will be downloaded at runtime from NatML Hub and cached onto the device, reducing the app size significantly. ***NOTE*** The Silero speech-to-text ML model is licensed under the GNU Affero GPL version 3 copyleft license. Ensure that your use case is compatible with this license. Read more. *********** Find other ML models on NatML Hub. Join the NatSuite community on Discord. Check out the NatML documentation. See more NatSuite projects on GitHub. Read the NatSuite blog. Discuss NatML on Unity Forums. Contact us at hi@natsuite.io.