Nuance Speech Recognition (SR) Plugin 1.4.0 to the UniMRCP Server (UMS) has been released.
The plugin is based on the following components:
- Nuance APIs 1.1.0
- gRPC 1.30.3
- Protobuf 3.12.2
- Libevent 2.1.9
- Rapidjson 1.1.0
The plugin supports the following Nuance Mix APIs:
- ASRaaS API v1
- NLUaaS API v1
- NRaaS API v1
The plugin is available for the following UMS versions:
The binaries are currently available for the following Linux distributions:
- Red Hat / CentOS 7
- unimrcp-nuance-sr-1.7.6-1.el7.x86_64.rpm
- unimrcp-nuance-sr-1.8.6-1.el7.x86_64.rpm
- Red Hat / CentOS 8
- unimrcp-nuance-sr-1.7.6-1.el8.x86_64.rpm
- unimrcp-nuance-sr-1.8.6-1.el8.x86_64.rpm
- Ubuntu 20.04 LTS
- unimrcp-nuance-sr_1.7.6-focal_amd64.deb
- unimrcp-nuance-sr_1.8.6-focal_amd64.deb
- Ubuntu 22.04 LTS
- unimrcp-nuance-sr_1.7.6-jammy_amd64.deb
- unimrcp-nuance-sr_1.8.6-jammy_amd64.deb
This release adds support for the DTMFRecognize method in the NRaaS API. The release provides minor fixes and improvements in the use of the NRaaS and ASRaaS APIs.
The detailed list of changes introduced in this release follows.
New Features
- Added support for the DTMFRecognize method in the NRaaS API.
- Added a new parameter grammar-mode to HTTP grammars used with the NRaaS API. The parameter can be set to either 'speech' (default) or 'dtmf' and indicates whether the HTTP grammar is for speech or DTMFs.
Fixed Problems
- Fixed a segfault encountered while processing the START-INPUT-TIMERS request when only a DTMF grammar is activated with the ASRaaS API.
- Decoupled the attribute vad-mode from the header field Sensitivity-Level. The attribute vad-mode can be set separately via vendor-specific parameters.
- Decoupled the speech-incomplete timeout used by the internal VAD detector from the timeout used by the NRaaS API.
Configuration Parameters
Miscellaneous
- Updated the Usage Guide to reflect the changes introduced in this release.
Visit the Nuance SR plugin page for more information.
https://unispeech.io/nuancesr
https://www.unimrcp.org/nuancesr
Thank you for using UniMRCP.
--
Arsen Chaloyan
Author of UniMRCP
http://www.unimrcp.org