ZoomRx Audio

ZoomRx Audio is a PHP wrapper package for audio manipulation and Speech-To-Text operations.

SpeechToText provides a simple and convenient way to utilize the Speech-to-Text services offered by Google, OpenAI, and AssemblyAI. This package serves as a wrapper for the respective APIs, allowing developers to easily transcribe speech audio files into text in their PHP applications.

Features

• Supports three major Speech-to-Text service providers: Google, OpenAI, and AssemblyAI.
• Offers customizable settings for transcribing audio files, such as language,
• Provides error handling and response parsing for easy integration into PHP applications.

Requirements

• PHP 7.4 or above
• cURL extension enabled
• API key or credentials for the respective Speech-to-Text service providers (Google, OpenAI-Whisper, and AssemblyAI)
• Python for audio manipulation
• Python pydub module
• Python jiwer module
• ffmpeg CLI

Usage

To use SpeechToText-PHP in your PHP application, follow these steps:

1. Instantiate the SpeechToText class with the desired provider (Google, OpenAI, or AssemblyAI), and pass in the required authentication credentials:

   use ZoomRx\Audio\SpeechToText\SpeechToTextFactory;
   
   $speechToTextObj = SpeechToTextFactory::create(SpeechToTextFactory::ASSEMBLYAI);
   $speechToTextObj->setCredentials([
       'api_credential' => 'YOUR-API-KEY'
   ]);

2. Set configurations such as language, punctuate, etc,:

   $speechToTextObj->setConfigurations([
       'language' => 'en-US',
       'speaker_labels' => true,
       'speakers_expected' => 2,
       'word_time' => true,
       'punctuate' => true
   ]);

3. Invoke transcribe function along with the file path and get results:

   $speechToTextResult = $speechToTextObj->transcribe($infile);
   $transcription = $speechToTextResult->getTranscription();