SoundHound SS Plugin 1.1.0 Released

[Arsen Chaloyan]

SoundHound Speech Synthesis (SS) Plugin 1.1.0 to the UniMRCP Server (UMS) has been released.

The plugin is based on the following components:

• UniMRCP Server 1.7.0
• Houndify SDK 1.0.0

The binaries are currently available for the following Linux distributions:

• Red Hat / CentOS 7 (unimrcp-soundhound-ss-1.7.1-1.el7.x86_64.rpm)
• Red Hat / CentOS 8 (unimrcp-soundhound-ss-1.7.1-1.el7.x86_64.rpm)
• Ubuntu 18.04 LTS (unimrcp-soundhound-ss_1.7.1-bionic_amd64.deb)
• Ubuntu 20.04 LTS (unimrcp-soundhound-ss_1.7.1-focal_amd64.deb)

This release sets the language tag in request submitted to SoundHound API and also implements an alternate method of speech synthesis.

The detailed list of changes introduced in this release follows.

New Features

• Transparently set the language tag in the requests submitted to SoundHound API.
• Added support for fetching synthesized audio data by the URL returned in response to submitted TTS request,which allows to start RTP streaming earlier for requests having larger TTS content. This behavior is disabled by default.

Fixed Problems

• None.

Configuration Parameters

• Changed the default language from 'en-US' to 'en' to conform to the format accepted by SoundHound API.
• Added a new attribute 'audio-as-url' to the element 'synth-settings'.

Miscellaneous

• Updated the Usage Guide to reflect the changes introduced in this release.

Visit the SoundHound SS plugin page for more information.

• https://unispeech.io/shss
  
• https://www.unimrcp.org/shss
  

Thank you for using UniMRCP.

--

Arsen Chaloyan
Author of UniMRCP
http://www.unimrcp.org