On-device speech-to-index engine powered by deep learning.

Overview

Octopus

GitHub

PyPI Maven Central Cocoapods

Made in Vancouver, Canada by Picovoice

Twitter URL YouTube Channel Views

Octopus is Picovoice's Speech-to-Index engine. It directly indexes speech without relying on a text representation. This acoustic-only approach boosts accuracy by removing out-of-vocabulary limitation and eliminating the problem of competing hypothesis (e.g. homophones)

Table of Contents

Demos

Python Demos

Install the demo package:

sudo pip3 install pvoctopusdemo

Run the following in the terminal:

octopus_demo  --access_key {AccessKey} --audio_paths ${AUDIO_PATHS}

Replace ${AccessKey} with your AccessKey obtained from Picovoice Console and ${AUDIO_PATHS} with a space-separated list of audio files. Octopus starts processing the audio files and asks you for search phrases and shows results interactively.

For more information about the Python demos go to demo/python.

C Demos

Build the demo:

cmake -S demo/c/ -B demo/c/build && cmake --build demo/c/build

Index a given audio file:

./demo/c/build/octopus_index_demo ${LIBRARY_PATH} ${ACCESS_KEY} ${AUDIO_PATH} ${INDEX_PATH}

Then search the index for a given phrase:

./demo/c/build/octopus_search_demo ${LIBRARY_PATH} ${MODEL_PATH} ${ACCESS_KEY} ${INDEX_PATH} ${SEARCH_PHRASE}

Replace ${LIBRARY_PATH} with path to appropriate library available under lib, ${ACCESS_KEY} with AccessKey obtained from Picovoice Console, ${AUDIO_PATH} with the path to a given audio file and format, ${INDEX_PATH} with the path to cached index file and ${SEARCH_PHRASE} to a search phrase.

For more information about C demos go to demo/c.

Android Demos

Using Android Studio, open demo/android/OctopusDemo as an Android project.

Replace "${YOUR_ACCESS_KEY_HERE}" inside MainActivity.java with your AccessKey obtained from Picovoice Console. Then run the demo.

For more information about Android demos go to demo/android.

iOS Demos

From the demo/ios/OctopusDemo, run the following to install the Octopus CocoaPod:

pod install

Replace "{YOUR_ACCESS_KEY_HERE}" inside ViewModel.swift with your AccessKey obtained from Picovoice Console. Then, using Xcode, open the generated OctopusDemo.xcworkspace and run the application.

For more information about iOS demos go to demo/ios.

Web Demos

From demo/web run the following in the terminal:

yarn
yarn start

(or)

npm install
npm run start

Open http://localhost:5000 in your browser to try the demo.

SDKs

Python

Create an instance of the engine:

import pvoctopus
access_key = ""  # AccessKey provided by Picovoice Console (https://picovoice.ai/console/)
handle = pvoctopus.create(access_key=access_key)

Index your raw audio data or file:

audio_data = [..]
metadata = handle.index(audio_data)
# or 
audio_file_path = "/path/to/my/audiofile.wav"
metadata = handle.index_file(audio_file_path)

Then search the metadata for phrases:

{match.end_sec} ({match.probablity})") ">
avocado_matches = matches['avocado']
for match in avocado_matches:
    print(f"Match for `avocado`: {match.start_sec} -> {match.end_sec} ({match.probablity})")

When done the handle resources have to be released explicitly:

handle.delete()

C

pv_octopus.h header file contains relevant information. Build an instance of the object:

    const char *model_path = "..."; // absolute path to the model file available at `lib/common/octopus_params.pv`
    const char *access_key = "..." // AccessKey provided by Picovoice Console (https://picovoice.ai/console/)
    pv_octopus_t *handle = NULL;
    pv_status_t status = pv_octopus_init(access_key, model_path, &handle);
    if (status != PV_STATUS_SUCCESS) {
        // error handling logic
    }

Index audio data using constructed object:

const char *audio_path = "..."; // absolute path to the audio file to be indexed
void *indices = NULL;
int32_t num_indices_bytes = 0;
pv_status_t status = pv_octopus_index_file(handle, audio_path, &indices, &num_indices_bytes);
if (status != PV_STATUS_SUCCESS) {
    // error handling logic
}

Search the indexed data:

const char *phrase = "...";
pv_octopus_match_t *matches = NULL;
int32_t num_matches = 0;
pv_status_t status = pv_octopus_search(handle, indices, num_indices_bytes, phrase, &matches, &num_matches);
if (status != PV_STATUS_SUCCESS) {
    // error handling logic
}

When done be sure to release the acquired resources:

pv_octopus_delete(handle);

Android

Create an instance of the engine:

import ai.picovoice.octopus.*;

final String accessKey = "..."; // AccessKey provided by Picovoice Console (https://picovoice.ai/console/)
try {
    Octopus handle = new Octopus.Builder(accessKey).build(appContext);
} catch (OctopusException ex) { }

Index audio data using constructed object:

final String audioFilePath = "/path/to/my/audiofile.wav"
try {
    OctopusMetadata metadata = handle.indexAudioFile(audioFilePath);
} catch (OctopusException ex) { }

Search the indexed data:

HashMap <String, OctopusMatch[]> matches = handle.search(metadata, phrases);

for (Map.Entry<String, OctopusMatch[]> entry : map.entrySet()) {
    final String phrase = entry.getKey();
    for (OctopusMatch phraseMatch : entry.getValue()){
        final float startSec = phraseMatch.getStartSec();
        final float endSec = phraseMatch.getEndSec();
        final float probability = phraseMatch.getProbability();
    }
}

When done be sure to release the acquired resources:

metadata.delete();
handle.delete();

iOS

Create an instance of the engine:

import Octopus

let accessKey : String = // .. AccessKey provided by Picovoice Console (https://picovoice.ai/console/)
do {
    let handle = try Octopus(accessKey: accessKey)
} catch { }

Index audio data using constructed object:

let audioFilePath = "/path/to/my/audiofile.wav"
do {
    let metadata = try handle.indexAudioFile(path: audioFilePath)
} catch { }

Search the indexed data:

let matches: Dictionary<String, [OctopusMatch]> = try octopus.search(metadata: metadata, phrases: phrases)
for (phrase, phraseMatches) in matches {
    for phraseMatch in phraseMatches {
        var startSec = phraseMatch.startSec;
        var endSec = phraseMatch.endSec;
        var probability = phraseMatch.probability;
    }
}

When done be sure to release the acquired resources:

handle.delete();

Web

Octopus is available on modern web browsers (i.e., not Internet Explorer) via WebAssembly. Octopus is provided pre-packaged as a Web Worker to allow it to perform processing off the main thread.

Vanilla JavaScript and HTML (CDN Script Tag)

">
>
<html lang="en">

<head>
  <script src="https://unpkg.com/@picovoice/octopus-web-en-worker/dist/iife/index.js">script>
  <script type="application/javascript">
    // The metadata object to save the result of indexing for later searches
    let octopusMetadata = undefined

    function octopusIndexCallback(metadata) {
      octopusMetadata = metadata
    }

    function octopusSearchCallback(matches) {
      console.log(`Search results (${matches.length}):`)
      console.log(`Start: ${match.startSec}s -> End: ${match.endSec}s (Probability: ${match.probability})`)
    }

    async function startOctopus() {
      // Create an Octopus Worker
      // Note: you receive a Worker object, _not_ an individual Octopus instance
      const accessKey = ... // AccessKey string provided by Picovoice Console (https://picovoice.ai/console/)
      const OctopusWorker = await OctopusWorkerFactory.create(
        accessKey,
        octopusIndexCallback,
        octopusSearchCallback
      )
    }

    document.addEventListener("DOMContentLoaded", function () {
      startOctopus();
      // Send Octopus the audio signal
      const audioSignal = new Int16Array(/* Provide data with correct format*/)
      OctopusWorker.postMessage({
        command: "index",
        input: audioSignal,
      });
    });

    const searchText = ...
    OctopusWorker.postMessage({
      command: "search",
      metadata: octopusMetadata,
      searchPhrase: searchText,
    });
  script>
head>

<body>body>

html>

Vanilla JavaScript and HTML (ES Modules)

yarn add @picovoice/octopus-web-en-worker

(or)

npm install @picovoice/octopus-web-en-worker
End: ${match.endSec}s (Probability: ${match.probability})`); } async function startOctopus() { // Create an Octopus Worker // Note: you receive a Worker object, _not_ an individual Octopus instance const accessKey = // .. AccessKey provided by Picovoice Console (https://picovoice.ai/console/) const OctopusWorker = await OctopusWorkerFactory.create( accessKey, octopusIndexCallback, octopusSearchCallback ); } startOctopus() ... // Send Octopus the audio signal const audioSignal = new Int16Array(/* Provide data with correct format*/) OctopusWorker.postMessage({ command: "index", input: audioSignal, }); ... const searchText = ...; OctopusWorker.postMessage({ command: "search", metadata: octopusMetadata, searchPhrase: searchText, }); ">
import { OctopusWebEnWorker } from "@picovoice/octopus-web-en-worker";

// The metadata object to save the result of indexing for later searches
let octopusMetadata = undefined;

function octopusIndexCallback(metadata) {
  octopusMetadata = metadata;
}

function octopusSearchCallback(matches) {
  console.log(`Search results (${matches.length}):`);
  console.log(`Start: ${match.startSec}s -> End: ${match.endSec}s (Probability: ${match.probability})`);
}


async function startOctopus() {
  // Create an Octopus Worker
  // Note: you receive a Worker object, _not_ an individual Octopus instance
  const accessKey = // .. AccessKey provided by Picovoice Console (https://picovoice.ai/console/)
  const OctopusWorker = await OctopusWorkerFactory.create(
    accessKey,
    octopusIndexCallback,
    octopusSearchCallback
  );
}

startOctopus()

...

// Send Octopus the audio signal
const audioSignal = new Int16Array(/* Provide data with correct format*/)
OctopusWorker.postMessage({
  command: "index",
  input: audioSignal,
});

...

const searchText = ...;
OctopusWorker.postMessage({
  command: "search",
  metadata: octopusMetadata,
  searchPhrase: searchText,
});

Releases

v1.0.0 Oct 8th, 2021

  • Initial release.
You might also like...
A python script to lookup Passport Index Dataset

visa-cli A python script to lookup Passport Index Dataset Installation pip install visa-cli Usage usage: visa-cli [-h] [-d DESTINATION_COUNTRY] [-f]

This is a virtual picture dragging application. Users may virtually slide photos across the screen. The distance between the index and middle fingers determines the movement. Smaller distances indicate click and motion, whereas bigger distances indicate only hand movement.
A set of simple scripts to process the Imagenet-1K dataset as TFRecords and make index files for NVIDIA DALI.

Overview This is a set of simple scripts to process the Imagenet-1K dataset as TFRecords and make index files for NVIDIA DALI. Make TFRecords To run t

PPLNN is a Primitive Library for Neural Network is a high-performance deep-learning inference engine for efficient AI inferencing
PPLNN is a Primitive Library for Neural Network is a high-performance deep-learning inference engine for efficient AI inferencing

PPLNN is a Primitive Library for Neural Network is a high-performance deep-learning inference engine for efficient AI inferencing

Implementation of "A Deep Learning Loss Function based on Auditory Power Compression for Speech Enhancement" by pytorch

This repository is used to suspend the results of our paper "A Deep Learning Loss Function based on Auditory Power Compression for Speech Enhancement"

Implementation of "StrengthNet: Deep Learning-based Emotion Strength Assessment for Emotional Speech Synthesis"

StrengthNet Implementation of "StrengthNet: Deep Learning-based Emotion Strength Assessment for Emotional Speech Synthesis" https://arxiv.org/abs/2110

This is the implementation of "SELF SUPERVISED REPRESENTATION LEARNING WITH DEEP CLUSTERING FOR ACOUSTIC UNIT DISCOVERY FROM RAW SPEECH" submitted to ICASSP 2022

CPC_DeepCluster This is the implementation of "SELF SUPERVISED REPRESENTATION LEARNING WITH DEEP CLUSTERING FOR ACOUSTIC UNIT DISCOVERY FROM RAW SPEEC

A fast, dataset-agnostic, deep visual search engine for digital art history

imgs.ai imgs.ai is a fast, dataset-agnostic, deep visual search engine for digital art history based on neural network embeddings. It utilizes modern

Comments
  • Bump terser from 5.13.1 to 5.16.1 in /binding/web

    Bump terser from 5.13.1 to 5.16.1 in /binding/web

    Bumps terser from 5.13.1 to 5.16.1.

    Changelog

    Sourced from terser's changelog.

    v5.16.1

    • Properly handle references in destructurings (const { [reference]: val } = ...)
    • Allow parsing of .#privatefield in nested classes
    • Do not evaluate operations that return large strings if that would make the output code larger
    • Make collapse_vars handle block scope correctly
    • Internal improvements: Typos (#1311), more tests, small-scale refactoring

    v5.16.0

    • Disallow private fields in object bodies (#1011)
    • Parse #privatefield in object (#1279)
    • Compress #privatefield in object

    v5.15.1

    • Fixed missing parentheses around optional chains
    • Avoid bare let or const as the bodies of if statements (#1253)
    • Small internal fixes (#1271)
    • Avoid inlining a class twice and creating two equivalent but !== classes.

    v5.15.0

    • Basic support for ES2022 class static initializer blocks.
    • Add AudioWorkletNode constructor options to domprops list (#1230)
    • Make identity function inliner not inline id(...expandedArgs)

    v5.14.2

    • Security fix for RegExps that should not be evaluated (regexp DDOS)
    • Source maps improvements (#1211)
    • Performance improvements in long property access evaluation (#1213)

    v5.14.1

    • keep_numbers option added to TypeScript defs (#1208)
    • Fixed parsing of nested template strings (#1204)

    v5.14.0

    • Switched to @​jridgewell/source-map for sourcemap generation (#1190, #1181)
    • Fixed source maps with non-terminated segments (#1106)
    • Enabled typescript types to be imported from the package (#1194)
    • Extra DOM props have been added (#1191)
    • Delete the AST while generating code, as a means to save RAM
    Commits

    Dependabot compatibility score

    Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.


    Dependabot commands and options

    You can trigger Dependabot actions by commenting on this PR:

    • @dependabot rebase will rebase this PR
    • @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
    • @dependabot merge will merge this PR after your CI passes on it
    • @dependabot squash and merge will squash and merge this PR after your CI passes on it
    • @dependabot cancel merge will cancel a previously requested merge and block automerging
    • @dependabot reopen will reopen this PR if it is closed
    • @dependabot close will close this PR and stop Dependabot recreating it. You can achieve the same result by closing it manually
    • @dependabot ignore this major version will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)
    • @dependabot ignore this minor version will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)
    • @dependabot ignore this dependency will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)
    • @dependabot use these labels will set the current labels as the default for future PRs for this repo and language
    • @dependabot use these reviewers will set the current reviewers as the default for future PRs for this repo and language
    • @dependabot use these assignees will set the current assignees as the default for future PRs for this repo and language
    • @dependabot use this milestone will set the current milestone as the default for future PRs for this repo and language

    You can disable automated security fix PRs for this repo from the Security Alerts page.

    dependencies 
    opened by dependabot[bot] 1
  • Convert c owned memory into ctypes owned memory for metadata objects

    Convert c owned memory into ctypes owned memory for metadata objects

    ctypes doesn't free the c_void_p from c, but if we convert it into a block of memory created from ctypes (serialize -> deserialize), then it will garbage collect and free when appropriate.

    opened by ErisMik 1
Releases(v1.2)
  • v1.2(Aug 12, 2022)

Owner
Picovoice
Edge Voice AI Platform
Picovoice
The implementation of "Shuffle Transformer: Rethinking Spatial Shuffle for Vision Transformer"

Shuffle Transformer The implementation of "Shuffle Transformer: Rethinking Spatial Shuffle for Vision Transformer" Introduction Very recently, window-

87 Nov 29, 2022
Multi-Content GAN for Few-Shot Font Style Transfer at CVPR 2018

MC-GAN in PyTorch This is the implementation of the Multi-Content GAN for Few-Shot Font Style Transfer. The code was written by Samaneh Azadi. If you

Samaneh Azadi 422 Dec 04, 2022
A Python implementation of the Locality Preserving Matching (LPM) method for pruning outliers in image matching.

LPM_Python A Python implementation of the Locality Preserving Matching (LPM) method for pruning outliers in image matching. The code is established ac

AoxiangFan 11 Nov 07, 2022
IDA file loader for UF2, created for the DEFCON 29 hardware badge

UF2 Loader for IDA The DEFCON 29 badge uses the UF2 bootloader, which conveniently allows you to dump and flash the firmware over USB as a mass storag

Kevin Colley 6 Feb 08, 2022
The official PyTorch implementation of paper BBN: Bilateral-Branch Network with Cumulative Learning for Long-Tailed Visual Recognition

BBN: Bilateral-Branch Network with Cumulative Learning for Long-Tailed Visual Recognition Boyan Zhou, Quan Cui, Xiu-Shen Wei*, Zhao-Min Chen This repo

Megvii-Nanjing 616 Dec 21, 2022
PyTorch implementation of Munchausen Reinforcement Learning based on DQN and SAC. Handles discrete and continuous action spaces

Exploring Munchausen Reinforcement Learning This is the project repository of my team in the "Advanced Deep Learning for Robotics" course at TUM. Our

Mohamed Amine Ketata 10 Mar 10, 2022
N-Omniglot is a large neuromorphic few-shot learning dataset

N-Omniglot [Paper] || [Dataset] N-Omniglot is a large neuromorphic few-shot learning dataset. It reconstructs strokes of Omniglot as videos and uses D

11 Dec 05, 2022
🤗 Transformers: State-of-the-art Natural Language Processing for Pytorch, TensorFlow, and JAX.

English | 简体中文 | 繁體中文 State-of-the-art Natural Language Processing for Jax, PyTorch and TensorFlow 🤗 Transformers provides thousands of pretrained mo

Hugging Face 77.2k Jan 02, 2023
Image-Scaling Attacks and Defenses

Image-Scaling Attacks & Defenses This repository belongs to our publication: Erwin Quiring, David Klein, Daniel Arp, Martin Johns and Konrad Rieck. Ad

Erwin Quiring 163 Nov 21, 2022
Fast Neural Representations for Direct Volume Rendering

Fast Neural Representations for Direct Volume Rendering Sebastian Weiss, Philipp Hermüller, Rüdiger Westermann This repository contains the code and s

Sebastian Weiss 20 Dec 03, 2022
Leaf: Multiple-Choice Question Generation

Leaf: Multiple-Choice Question Generation Easy to use and understand multiple-choice question generation algorithm using T5 Transformers. The applicat

Kristiyan Vachev 62 Dec 20, 2022
TeST: Temporal-Stable Thresholding for Semi-supervised Learning

TeST: Temporal-Stable Thresholding for Semi-supervised Learning TeST Illustration Semi-supervised learning (SSL) offers an effective method for large-

Xiong Weiyu 1 Jul 14, 2022
A Broader Picture of Random-walk Based Graph Embedding

Random-walk Embedding Framework This repository is a reference implementation of the random-walk embedding framework as described in the paper: A Broa

Zexi Huang 23 Dec 13, 2022
Learning an Adaptive Meta Model-Generator for Incrementally Updating Recommender Systems

Learning an Adaptive Meta Model-Generator for Incrementally Updating Recommender Systems This is our experimental code for RecSys 2021 paper "Learning

11 Jul 28, 2022
Code for "Neural 3D Scene Reconstruction with the Manhattan-world Assumption" CVPR 2022 Oral

News 05/10/2022 To make the comparison on ScanNet easier, we provide all quantitative and qualitative results of baselines here, including COLMAP, COL

ZJU3DV 365 Dec 30, 2022
Convert human motion from video to .bvh

video_to_bvh Convert human motion from video to .bvh with Google Colab Usage 1. Open video_to_bvh.ipynb in Google Colab Go to https://colab.research.g

Dene 306 Dec 10, 2022
StorSeismic: An approach to pre-train a neural network to store seismic data features

StorSeismic: An approach to pre-train a neural network to store seismic data features This repository contains codes and resources to reproduce experi

Seismic Wave Analysis Group 11 Dec 05, 2022
A Real-Time-Strategy game for Deep Learning research

Description DeepRTS is a high-performance Real-TIme strategy game for Reinforcement Learning research. It is written in C++ for performance, but provi

Centre for Artificial Intelligence Research (CAIR) 156 Dec 19, 2022
The spiritual successor to knockknock for PyTorch Lightning, get notified when your training ends

Who's there? The spiritual successor to knockknock for PyTorch Lightning, to get a notification when your training is complete or when it crashes duri

twsl 70 Oct 06, 2022
NAS-HPO-Bench-II is the first benchmark dataset for joint optimization of CNN and training HPs.

NAS-HPO-Bench-II API Overview NAS-HPO-Bench-II is the first benchmark dataset for joint optimization of CNN and training HPs. It helps a fair and low-

yoichi hirose 8 Nov 21, 2022