Installation Guide (PSP)
This article describes how to perform basic installation of Speech Platform (i.e. Speech Engine component and Phonexia Browser component).
Prerequisites
For further information we strongly recommend checking our SPE Training videos, especially “Introduction” and “Module 1”.
In order to install Phonexia Speech Platform (PSP) you need to download it first.
You can download installation package on our partner portal, using this link:
https://partner.phonexia.com/kb/sp/speech-platform/evaluation-package/
Package content – Installation package contains following components, each in separate directory.
- Speech Engine (SPE) is the core of Phonexia speech platform. It is backend application that performs all the work. It processes files and returns desired result from each speech technology (age estimation, transcript, speaker identification and others). Communication with SPE is handled exclusively via REST API, it does not have own GUI.
- Phonexia Browser (Browser) is graphical user interface client that communicates with SPE via REST API. It is used mostly for testing, demonstration, and evaluation of results. Phonexia Browser does not process any files itself, all requests are processed by SPE, which can run either standalone (even on remote location) or embedded.
Note: do not confuse Phonexia Browser with web browser.
Types of installation
There are three main types of installation of Speech Platform.
Embedded SPE to Browser
This is the most basic and easiest option, used by default in the Evaluation Package. You just run Phonexia Browser, which starts SPE in the background.
You can then control basic SPE configuration (technologies setup, certain SPE settings, SPE restarts) directly from within Browser.
|
It is most suitable for:
|
It is not suitable for:
|
Standalone SPE with Browser (recommended)
In this setup is SPE run as separate standalone process and Browser is also started separately. Both can be run on the same machine, or on separate machines as typical client–server configuration. You can then connect to SPE either using Browser, or with other application using the SPE REST API.
Note that in this setup it is not possible to control SPE configuration (technologies setup, SPE settings, SPE restarts) from within Browser, it must be done separately using appropriate tools/methods.
|
It is mostly suitable for:
|
It is not suitable for:
|
Standalone SPE
It is a variant of the previous setup, suitable for heavy load. This option doesn’t incorporate Browser or any other GUI tool (surely, Browser can be still connected to SPE on client–server basis any time). Communication with SPE is handled purely via REST API.
This option is very versatile based on your needs and other variables. It is discussed in other articles or can be explained further by Phonexia technical consultant.
|
It is mostly suitable for:
|
It is not suitable for:
|
Installation
IMPORTANT: This article covers fast local installation for testing, demonstration, and development purposes. To install Speech Platform on server for production use, follow information in bundled SPE/doc/install.htm document.
Installation is very simple. You downloaded ZIP file. Entire package is portable, so you can just unzip the package into any folder that you need.
We recommend creating C:\phonexia (on Windows) or $HOME/phonexia (on Linux) directory and extract the ZIP archive there.
You should also obtain license.dat file. It is not strictly required for installation, but you will need to start and run Phonexia software. You can request the license file from your Phonexia contact.
If you have installed Speech Platform successfully, we recommend checking Quick Start Guide (PSP). You might be also interested in How to install new language or technology model.