Picovoice Platform — Java Quick Start

End-to-End Voice Platform
Offline Voice Recognition
Local Speech Recognition
Speech-to-Intent
Domain-Specific NLU
Wake Word Detection
Linux
macOS
Windows
Java

Requirements

Java 11+

Compatibility

Linux (x86_64)
macOS (x86_64)
Windows (x86_64)

Cloning the Picovoice Repository

If using SSH, clone the Picovoice repository with:

git clone --recurse-submodules [email protected]:Picovoice/picovoice.git

If using HTTPS, then type:

git clone --recurse-submodules https://github.com/Picovoice/picovoice.git

Installation

You can get the latest Java demo executable JARs here.

If you wish, you can build the demos from source by opening the project with the IntelliJ IDE. Select "Build > Build Project" to build the two demo classes or "Build > Build Artifacts" to create the executable JARs.

Usage

NOTE: the working directory for all dotnet commands is:

picovoice/demo/java/bin

File Demo

The file demo uses Picovoice to scan for keywords and commands in an audio file. The demo is mainly useful for quantitative performance benchmarking against a corpus of audio data.

Picovoice engines require 16-bit 16kHz linearly-encoded PCM (single-channel) audio.

Picovoice processes a 16kHz, single-channel audio stream. If a stereo file is provided it only processes the first (left) channel.

The following command processes a file looking for instances of the wake phrase defined in the file located at ${PATH_TO_PORCUPINE_KEYWORD_FILE} and then infers the follow-on spoken command using the context defined by the file located at ${PATH_TO_RHINO_CONTEXT_FILE)}:

java -jar picovoice-file-demo.jar \
-i ${PATH_TO_INPUT_AUDIO_FILE} \
-k ${PATH_TO_PORCUPINE_KEYWORD_FILE} \
-c ${PATH_TO_RHINO_CONTEXT_FILE}

Microphone Demo

This demo opens an audio stream from a microphone and detects utterances of a give wake word(s). The following processes incoming audio from the microphone for instances of the wake phrase defined in the file located at ${PATH_TO_PORCUPINE_KEYWORD_FILE} and then infers the follow-on spoken command using the context defined by the file located at ${PATH_TO_RHINO_CONTEXT_FILE)}. Upon completion of the spoken command inference it resumes wake word detection.

java -jar picovoice-mic-demo.jar \
-k ${PATH_TO_PORCUPINE_KEYWORD_FILE} \
-c ${PATH_TO_RHINO_CONTEXT_FILE}

It is possible that the default audio input device recognized by PyAudio is not the one being used. There are a couple of debugging facilities baked into the demo application to solve this. First, type the following into the console:

java -jar picovoice-mic-demo.jar -sd

It provides information about various audio input devices on the box. On a sample Windows PC, this is the output:

Available input devices:
 
    Device 0: Microphone Array (Realtek(R) Au
    Device 1: Microphone Headset USB

You can use the device index to specify which microphone to use for the demo. For instance, if you want to use the Headset microphone in the above example, you can invoke the demo application as below:

java -jar picovoice-mic-demo.jar \
-k ${PATH_TO_PORCUPINE_KEYWORD_FILE} \
-c ${PATH_TO_RHINO_CONTEXT_FILE}
-di 1

If the problem persists we suggest storing the recorded audio into a file for inspection. This can be achieved with:

java -jar picovoice-mic-demo.jar \
-k ${PATH_TO_PORCUPINE_KEYWORD_FILE} \
-c ${PATH_TO_RHINO_CONTEXT_FILE)} \
-di 1 \
-o ./test.wav

If after listening to stored file there is no apparent problem detected, please open an issue on GitHub.