Google Speech Synthesis (GSS) Plugin 1.24.0 to the UniMRCP Server (UMS) has been released.
The plugin is based on the following components:
- Google APIs 1.10.0
- gRPC 1.61.1
- Protobuf 25.1
The plugin supports the following Google APIs:
- Text-to-Speech v1
- GenerativeLanguage v1beta (Gemini Flash TTS)
The plugin is available for the following UMS versions:
The binaries are currently available for the following Linux distributions:
- Red Hat 8
- unimrcp-gss-1.7.23-1.el8.x86_64.rpm
- unimrcp-gss-1.8.10-1.el8.x86_64.rpm
- Red Hat 9
- unimrcp-gss-1.7.23-1.el9.x86_64.rpm
- unimrcp-gss-1.8.10-1.el9.x86_64.rpm
- Ubuntu 22.04 LTS
- unimrcp-gss_1.7.23-jammy_amd64.deb
- unimrcp-gss_1.8.10-jammy_amd64.deb
- Ubuntu 24.04 LTS
- unimrcp-gss_1.7.23-noble_amd64.deb
- unimrcp-gss_1.8.10-noble_amd64.deb
This
release adds support for flattening supplied SSML content into plain
text segments to be used with the method StreamingSynthesize. The
release allows to save original and resampled waveforms applicable to
the use of Gemini Flash voices.
The detailed list of changes introduced in this release follows.
New Features- Allow
to flatten received SSML content and pass plain text segments to the
service endpoint for speech synthesis applicable to the method
StreamingSynthesize.
- Allow to save original waveform before resampling is applied, applicable to Gemini Flash voices only.
Fixed Problems- Fixed processing of a special case when the method BidiGenerateContent completes successfully but with no synthesized audio data.
Configuration Parameters- Added a new attribute 'flatten-ssml' to the element 'synth-settings'.
- Added a new attribute 'save-orig-waveforms' to the element 'waveform-manager'.
Miscellaneous- Updated the Usage Guide to reflect the changes introduced in this release.
Visit the GSS plugin page for more information.
Thank you for using UniMRCP.