VDK 5.5.0 - 8th January 2024

Summary

We are pleased to introduce VDK 5.5.0 which is a feature release.

This version features an improved way of using the Text-to-Speech technology: the streaming.
It allows a better experience by starting to play generated audio before the end of the whole generation, on the fly!

Regarding the voice biometrics technology, the audio used for training can now be quality checked. For example you can now know how much speech has been extracted from the audio used specifically for training.

The studio has also seen its stability increased.
You can find more details about other minor changes described below, under relevant components.

Versions

Studio

Component	Version
Studio	5.8.7 UPDATED
Dictionary manager	2.1.3 UPDATED
Free-speech	5.2.1 UPDATED
Grammar editor	5.2.3 UPDATED
Phonetic editor	5.1.3 UPDATED
Project editor	2.3.1 UPDATED
Simple assistant maker	5.2.1 UPDATED
TTS manager	5.2.2 UPDATED
Voice biometrics	4.4.4 UPDATED
Nlu Editor	2.1.4 UPDATED
Speech Enhancer	1.1.1 UPDATED
Vsdk-daemon	1.14.9 UPDATED

Samples

Component	Version	Technologies
Chained grammars	5.1.2 UPDATED	VOICE RECOGNITION
Dynamic grammar	5.1.2 UPDATED	VOICE RECOGNITION
Simple application	5.1.1 UPDATED	VOICE RECOGNITION VOICE SYNTHESIS
Tts	5.1.1 UPDATED	VOICE SYNTHESIS
Voice biometrics	5.1.2 UPDATED	VOICE BIOMETRICS
Voice Commands Language Understanding	2.1.1 UPDATED	VOICE RECOGNITION
Speech Enhancement	1.0.1 UPDATED	VOICE RECOGNITION SPEECH ENHANCEMENT

VSDK (C++)

Component	Version
Vsdk	9.1.1 UPDATED
Vsdk-vasr	5.2.6 UPDATED
Vsdk-csdk	5.1.4 UPDATED
Vsdk-tnl	3.2.1 UPDATED
Vsdk-baratinoo	4.2.1 UPDATED
Vsdk-vtapi	3.2.2 UPDATED
Vsdk-idvoice	4.3.1 UPDATED
Vsdk-tssv	3.3.2 UPDATED
Vsdk-vnlu	2.2.1 UPDATED
Vsdk-audio-portaudio	3.5.2 UPDATED
Vsdk-s2c	1.1.2 UPDATED

VSDK (JAVA)

Component	Version
Vsdk	5.3.0
Vsdk-csdk	5.1.2 UPDATED
Vsdk-tnl	2.5.0
Vsdk-baratinoo	4.0.0
Vsdk-vtapi	3.1.0
Vsdk-idvoice	2.5.0
Vsdk-tssv	2.6.0
Vsdk-vasr	2.2.1 UPDATED
Vsdk-vnlu	1.0.2
Vsdk-s2c	1.1.0 UPDATED

Details

Studio

Bug-fixes:

Changing license was not correctly applying changes of permissions

Project Editor

Bug-fixes:

VDK project files could have wrong path separator when opened on another platform
Changing voice recognition’s SDK could sometimes lead to invalid selected SDK

Improvements:

Removed instance count in technology cards for an improved user experience
Limited access to technologies are better displayed

Text-to-Speech

Bug-fixes:

The Download more voices button now auto filter voices with the current SDK
Using dictionaries could sometimes have incorrect text escaping

Improvements:

Now streams generated audio for a better response time while using very long texts

Voice Biometrics

Improvements:

Adding audio for training now accounts for speech quality in the count of seconds or number of utterances

Bug-fixes:

Closing the studio while recording audio for training could lead to a crash

NLU Editor

Bug-fixes:

Parallel trainings could fail if started in a very specific short time frame
During a training, if the remote sends an error, the widget would simply reset its state to Not training without any notice
Upon adding an example, the focus would not get back to adding a new example

Simple Assistant Maker

Bug-fixes:

Assistant removal could, in some specific circumstances, lead to several assistants removal

Grammar Editor

Bug-fixes:

Really fast consecutive quick testing could trigger a Engine is busy error message

Voice Biometrics

Improvements:

Adding audio for training now accounts for speech quality in the count of seconds or number of utterances

Samples (C++)

Chained grammars

Improvements:

Now use a utility library vsdk-samples-utils to wrap the custom Event loop system

Dynamic grammar

Improvements:

Now use a utility library vsdk-samples-utils to wrap the custom Event loop system

Simple application

Bug-fixes:

When using Vsdk-vtapi as the TTS SDK, it could happen that the synthesized audio gets truncated

Improvements:

Now use a utility library vsdk-samples-utils to wrap the custom Event loop system

Tts

Improvements:

Now use a utility library vsdk-samples-utils to wrap the custom Event loop system

Voice Biometrics

Improvements:

Now use a utility library vsdk-samples-utils to wrap the custom Event loop system

Voice Commands Language Understanding

Improvements:

Now use a utility library vsdk-samples-utils to wrap the custom Event loop system

VSDK (C++)

Vsdk

Bug-fixes:

Upon error, a crash could happen on corner cases when the producer was invalid

Vsdk-vasr

Improvements:

Performances and accuracy have been significantly improved

Vsdk-tssv

Bug-fixes:

Could not analyze more than 12 seconds of audio

VSDK (Java)

Vsdk-vasr

Improvements:

Performances and accuracy have been significantly improved