Riva SS Plugin 1.2.0 Released

[Arsen Chaloyan]

Riva Speech Synthesis (SS) Plugin 1.2.0 to the UniMRCP Server (UMS) has been released.

The plugin is based on the following components:

• NVIDIA Riva Speech Skills 2.15.0
• gRPC 1.30.3
• Protobuf 3.12.2

The plugin is available for the following UMS versions:

• 1.7.0
• 1.8.0

The binaries are currently available for the following Linux distributions:

• Red Hat / CentOS 7
  • unimrcp-riva-ss-1.7.3-1.el7.x86_64.rpm
  • unimrcp-riva-ss-1.8.2-1.el7.x86_64.rpm
• Red Hat / CentOS 8
  • unimrcp-riva-ss-1.7.3-1.el8.x86_64.rpm
  • unimrcp-riva-ss-1.8.2-1.el8.x86_64.rpm
• Ubuntu 20.04 LTS
  • unimrcp-riva-ss_1.7.3-focal_amd64.deb
  • unimrcp-riva-ss_1.8.2-focal_amd64.deb
• Ubuntu 22.04 LTS
  • unimrcp-riva-ss_1.7.3-jammy_amd64.deb
  • unimrcp-riva-ss_1.8.2-jammy_amd64.deb

This release is built against the latest NVIDIA Riva Speech Skills 2.15.0. The release provides a few minor fixes and enhancements.

The detailed list of changes introduced in this release follows.

New Features

• None.

Fixed Problems

• Fixed processing of max-file-age, which can optionally be set for waveforms, to properly remove file entries when max-file-age is reached.
• Fixed a race condition issue in processing SPEAK -> STOP -> SPEAK requests. The problem is encountered when the first SPEAK request is placed via HTTP to the service endpoint and stopped immediately, while the second SPEAK request is served from the local cache.
• Make sure special characters are always escaped in SSML.

Configuration Parameters

• None.

Miscellaneous

• Updated the NVIDIA Riva Speech Skills from 2.0.0 to 2.15.0.

Visit the Riva SS plugin page for more information.

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

Thank you for using UniMRCP.

--

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