Cheetah Binding for Web

Cheetah Speech-to-Text Engine

Made in Vancouver, Canada by Picovoice

Cheetah is an on-device streaming speech-to-text engine. Cheetah is:

• Private; All voice processing runs locally.
• Accurate
• Compact and Computationally-Efficient
• Cross-Platform:
  • Linux (x86_64), macOS (x86_64, arm64), and Windows (x86_64)
  • Android and iOS
  • Chrome, Safari, Firefox, and Edge
  • Raspberry Pi (3, 4, 5)

Compatibility

• Chrome / Edge
• Firefox
• Safari

Restrictions

IndexedDB is required to use Cheetah in a worker thread. Browsers without IndexedDB support (i.e. Firefox Incognito Mode) should use Cheetah in the main thread.

Installation

Package

Using Yarn:

yarn add @picovoice/cheetah-web

or using npm:

npm install --save @picovoice/cheetah-web

AccessKey

Cheetah requires a valid Picovoice AccessKey at initialization. AccessKey acts as your credentials when using Cheetah SDKs. You can get your AccessKey for free. Make sure to keep your AccessKey secret. Signup or Login to Picovoice Console to get your AccessKey.

Usage

Create a model in Picovoice Console or use the default model.

For the web packages, there are two methods to initialize Cheetah.

Public Directory

NOTE: Due to modern browser limitations of using a file URL, this method does not work if used without hosting a server.

This method fetches the model file from the public directory and feeds it to Cheetah. Copy the model file into the public directory:

cp ${CHEETAH_MODEL_FILE} ${PATH_TO_PUBLIC_DIRECTORY}

Base64

NOTE: This method works without hosting a server, but increases the size of the model file roughly by 33%.

This method uses a base64 string of the model file and feeds it to Cheetah. Use the built-in script pvbase64 to base64 your model file:

npx pvbase64 -i ${CHEETAH_MODEL_FILE} -o ${OUTPUT_DIRECTORY}/${MODEL_NAME}.js

The output will be a js file which you can import into any file of your project. For detailed information about pvbase64, run:

npx pvbase64 -h

Cheetah Model

Cheetah saves and caches your model file in IndexedDB to be used by WebAssembly. Use a different customWritePath variable to hold multiple models and set the forceWrite value to true to force re-save a model file.

Either base64 or publicPath must be set to instantiate Cheetah. If both are set, Cheetah will use the base64 model.

const cheetahModel = {
  publicPath: ${MODEL_RELATIVE_PATH},
  // or
  base64: ${MODEL_BASE64_STRING},

  // Optionals
  customWritePath: "cheetah_model",
  forceWrite: false,
  version: 1,
}

Init options

Set endpointDurationSec value to 0 if you do not wish to detect endpoint (moment of silence). Set enableAutomaticPunctuation to true to enable punctuation in transcript. Set processErrorCallback to handle errors if an error occurs while transcribing.

// Optional, these are default
const options = {
  endpointDurationSec: 1.0,
  enableAutomaticPunctuation: false,
  processErrorCallback: (error) => {}
}

Initialize Cheetah

Create a transcriptCallback function to get the streaming results from the engine:

let transcript = "";

function transcriptCallback(cheetahTranscript: CheetahTranscript) {
  transcript += cheetahTranscript.transcript;
  if (cheetahTranscript.isEndpoint) {
    transcript += ". ";
  }
  if (cheetahTranscript.isFlushed) {
    transcript += "\n"
  }
}

Create an instance of Cheetah on the main thread:

const handle = await Cheetah.create(
  ${ACCESS_KEY},
  transcriptCallback,
  cheetahModel,
  options // optional options
);

Or create an instance of Cheetah in a worker thread:

const handle = await CheetahWorker.create(
  ${ACCESS_KEY},
  transcriptCallback,
  cheetahModel,
  options // optional options
);

Process Audio Frames

The process function will send the input frames to the engine. The transcript is received from transcriptCallback as mentioned above.

function getAudioData(): Int16Array {
  ... // function to get audio data
  return new Int16Array();
}

for (;;) {
  handle.process(getAudioData());
  // break on some condition
}
handle.flush(); // runs transcriptCallback on remaining data.

Clean Up

Clean up used resources by Cheetah or CheetahWorker:

await handle.release();

Terminate (Worker only)

Terminate CheetahWorker instance:

await handle.terminate();

Language Model

Default models for supported languages can be found in lib/common.

Create custom language models using the Picovoice Console. Here you can train language models with custom vocabulary and boost words in the existing vocabulary.

Demo

For example usage refer to our Web demo application.