Eagle Binding for Web

Eagle Speaker Recognition Engine

Made in Vancouver, Canada by Picovoice

Eagle is an on-device speaker recognition engine. Eagle is:

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

Compatibility

• Chrome / Edge
• Firefox
• Safari

Requirements

The Eagle Web Binding uses SharedArrayBuffer.

Include the following headers in the response to enable the use of SharedArrayBuffers:

Cross-Origin-Opener-Policy: same-origin
Cross-Origin-Embedder-Policy: require-corp

Refer to our Web demo for an example on creating a server with the corresponding response headers.

Browsers that don't support SharedArrayBuffers or applications that don't include the required headers will fall back to using standard ArrayBuffers. This will disable multithreaded processing.

Restrictions

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

Multi-threading is only enabled for Eagle when using on a web worker.

Installation

Using yarn:

yarn add @picovoice/eagle-web

or using npm:

npm install --save @picovoice/eagle-web

AccessKey

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

Eagle has two distinct steps: Enrollment and Recognition. In the enrollment step, Eagle analyzes a series of utterances from a particular speaker to learn their unique voiceprint. This step produces a profile object, which can be stored and utilized during inference. During the Recognition step, Eagle compares the incoming frames of audio to the voiceprints of all enrolled speakers in real-time to determine the similarity between them.

Speaker Enrollment

Create an instance of the EagleProfiler:

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

// Main thread
const eagleProfiler = await EagleProfiler.create(
        ${ACCESS_KEY},
        eagleModel);

// or on worker thread
const eagleProfiler = await EagleProfilerWorker.create(
        ${ACCESS_KEY},
        eagleModel);

EagleProfiler is responsible for processing and enrolling PCM audio data, with the valid audio sample rate determined by eagleProfiler.sampleRate. The audio data must be 16-bit linearly-encoded and single-channel.

When passing samples to eagleProfiler.enroll, the number of samples must be equal to eagleProfiler.frameLength. The percentage value returned from this process indicates the progress of enrollment.

function getAudioData(numSamples): Int16Array {
  // get audio frame of size `numSamples`
}

let percentage = 0;
while (percentage < 100) {
  const audioData = getAudioData(eagleProfiler.frameLength);

  percentage = EagleProfilerEnrollResult = await eagleProfiler.enroll(audioData);
}

After the percentage reaches 100%, the enrollment process is considered complete. While it is possible to continue providing additional audio data to the profiler to improve the accuracy of the voiceprint, it is not necessary to do so.

const speakerProfile: EagleProfile = eagleProfiler.export();

The eagleProfiler.export() function produces a binary array, which can be saved to a file.

To reset the profiler and enroll a new speaker, the eagleProfiler.reset() method can be used. This method clears all previously stored data, making it possible to start a new enrollment session with a different speaker.

Finally, when done be sure to explicitly release the resources:

eagleProfiler.release();

// if on worker thread
eagleProfiler.terminate();

Speaker Recognition

Create an instance of the engine:

// Main thread
const eagle = await Eagle.create(
        ${ACCESS_KEY},
        eagleModel);

// or, on a worker thread
const eagle = await EagleWorker.create(
        ${ACCESS_KEY},
        eagleModel);

When initialized, eagle.sampleRate specifies the valid sample rate for Eagle. The expected length of a frame, or the minimum number of audio samples in an input array, is defined by eagle.minProcessSamples.

Like the profiler, Eagle is designed to work with single-channel audio that is encoded using 16-bit linear PCM.

Process the audio data with one or more speaker profiles created by the EagleProfiler:

function getAudioData(numSamples): Int16Array {
  // get audio frame of size `numSamples`
}

while (true) {
  const audioData = getAudioData(eagle.minProcessSamples);
  const scores: number[] | null = await eagle.process(
    audioData,
    speakerProfile
  );
  if (scores) {
    // do something with the scores
  }
}

The return value scores represents the degree of similarity between the input audio frame and the enrolled speakers. Each value is a floating-point number ranging from 0 to 1, with higher values indicating a greater degree of similarity. A return value of null indicates that there was not enough voice in the sample to detect a speaker.

Finally, when done be sure to explicitly release the resources:

eagle.release();

// if on worker thread
eagle.terminate();

Eagle Model

The default model is located in lib/common. Use it with the EagleModel type:

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

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

Eagle 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 Eagle. If both are set, Eagle will use the base64 model.

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 passes it to Eagle. Copy the model file into the 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 passes it to Eagle. Use the built-in script pvbase64 to base64 your model file:

npx pvbase64 -i ${EAGLE_MODEL_PATH} -o ${BASE64_MODEL_PATH}.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

Demos

For example usage refer to our Web demo application.