Leopard Binding for React

Leopard Speech-to-Text Engine

Made in Vancouver, Canada by Picovoice

Leopard is an on-device speech-to-text engine. Leopard is:

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

Compatibility

• Chrome / Edge
• Firefox
• Safari

Restrictions

IndexedDB and WebWorkers are required to use Leopard React. Browsers without support (e.g. Firefox Incognito Mode) should use the LeopardWeb binding main thread method.

AccessKey

Leopard requires a valid Picovoice AccessKey at initialization. AccessKey acts as your credentials when using Leopard 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.

Installation

Using yarn:

yarn add @picovoice/leopard-react @picovoice/web-voice-processor

or using npm:

npm install --save @picovoice/leopard-react @picovoice/web-voice-processor

Usage

Leopard requires a model file (.pv) at initialization. Use one of the default language models found in lib/common, or create a custom Leopard model in the Picovoice Console for the target platform Web (WASM).

There are two methods to initialize Leopard.

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 Leopard. Copy the model file into the public directory:

cp ${LEOPARD_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 Leopard. Use the built-in script pvbase64 to base64 your model file:

npx pvbase64 -i ${LEOPARD_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

Leopard Model

Leopard 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. If the model file changes, version should be incremented to force the cached models to be updated. Either base64 or publicPath must be set to instantiate Leopard. If both are set, Leopard will use the base64 model.

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

  // Optionals
  customWritePath: "custom_model",
  forceWrite: true,
  version: 1,
}

Additional engine options are provided via the options parameter. Set enableAutomaticPunctuation to true if you wish to enable punctuation in the transcript.

// Optional
const options = {
  enableAutomaticPunctuation: true
}

Initialize Leopard

Use useLeopard and init to initialize Leopard:

const {
  result,
  isLoaded,
  error,
  init,
  processFile,
  startRecording,
  stopRecording,
  isRecording,
  recordingElapsedSec,
  release,
} = useLeopard();

const initLeopard = async () => {
  await init(
    "${ACCESS_KEY}",
    leopardModel,
    options,
  );
}

In case of any errors, use the error state variable to check the error message. Use the isLoaded state variable to check if Leopard has loaded.

Transcribe Audio

The audio that you want to transcribe can either be uploaded as a File object or recorded with a microphone.

File Object

Transcribe File objects directly using the processFile function:

<input
  type="file"
  accept="audio/*"
  onChange={async (e) => {
    if (!!e.target.files?.length) {
      await processFile(e.target.files[0]);
    }
  }}
/>

Once the audio has been processed, the transcript will be available in the result state variable.

Record Audio

Leopard React binding uses WebVoiceProcessor to record audio. To start recording audio, call startRecording:

await startRecording();

If WebVoiceProcessor has started correctly, isRecording will be set to true.

Note: By default, Leopard will only record for 2 minutes before stopping and processing the buffered audio. This is to prevent unbounded memory usage. To increase this limit, call startRecording with the optional maxRecordingSec parameter:

const maxRecordingSec = 60 * 10 
await startRecording(maxRecordingSec)

Call stopRecording to stop recording audio:

await stopRecording();

If WebVoiceProcessor has stopped correctly, isRecording will be set to false. Once stopped, audio processing will automatically begin. Once completed, the transcript will be available in result.

Result

Once audio has been processed, the transcript will be available in the result state variable:

useEffect(() => {
  if (result !== null) {
    console.log(result.transcript);
    console.log(result.words);
  }
}, [result])

• transcript: A string containing the transcribed data.
• words: A list of objects containing a word, startSec, endSec, and confidence. Each object indicates the start, end time and confidence (between 0 and 1) of the word.

Release

While running in a component, you can call release to clean up all resources used by Leopard and WebVoiceProcessor:

await release();

This will set isLoaded and isRecording to false, recordingElapsedSec to 0, and error to null.

If any arguments require changes, call release, then init again to initialize Leopard with the new settings.

You do not need to call release when your component is unmounted - the hook will clean up automatically on unmount.

Non-English Languages

In order to transcribe non-English audio files and recordings you need to use the corresponding model file (.pv). The model files for all supported languages are available here.

Demo

For example usage refer to our React demo application.