Japan
Asked
Resolved Resolved by ptp!

Use Voice Recognition For Unsupported Languages

I am working on a science project and I have had good contact with problems but there is one problem I cannot do well - using languages that are not supported for speech recognition: I used downloaded language packs and Api.AI service with dialogflow or AZURE with SDK support but only works with text to speech conversion, but voice recognition is always understood by ARC. into English and cannot be used with native languages. I am a trainee in Vietnam and will need to use Vietnamese speech recognition for the next 5 months science and technology creative exam. I hope someone in the community will guide me before I quit useless time to create plugin for LUIS microsolf service? Thank you !!


Related Hardware EZ-B v4

ARC Pro

Upgrade to ARC Pro

Experience the transformation – subscribe to Synthiam ARC Pro and watch your robot evolve into a marvel of innovation and intelligence.

PRO
USA
#1  

@Joinny,

in a previous post https://synthiam.com/Community/Questions/Inconsistent-Speed-Speech-Settings-after-Installing-Cepstral-18209 (#8) you asked the same question regarding Korean Language.

Your profile's country is Japan

now you mentioned you are a trainee in Vietnam.

Please be coherent ... otherwise it seems you are jumping around asking superfluous questions.

Azure Speech to Text: https://docs.microsoft.com/en-us/azure/cognitive-services/speech-service/language-support

Languages supported (Speech Service): User-inserted image User-inserted image

Quote:

will need to use Vietnamese speech recognition for the next 5 months science
It's clear Azure Speech to text does not support Vietnamese language. It's not a plugin issue or a community help. If you are student you need to adapt your expectations to the existent technology, maybe there is a local company / university working on Vietnamese language, some local universities have government funds to develop local market solutions.

Quote:

I quit useless time to create plugin for LUIS microsolf service?
I don't understand the question. Microsoft LUIS is a different product/API basically is a text language parser, it works with text inputs (strings) extracts intentions and entities.

Microsoft Luis supported languages: https://docs.microsoft.com/en-us/azure/cognitive-services/luis/luis-language-support

User-inserted image Vietnamese is not supported BUT:

Quote:

If you need to provide language understanding for a language LUIS does not support, you can use Microsoft Translator API to translate the utterance into a supported language, submit the utterance to the LUIS endpoint, and receive the resulting scores.
Language and region support for the Translator Text API: https://docs.microsoft.com/en-us/azure/cognitive-services/translator/language-support User-inserted image

It seems there are some (not full) support for Vietnamese you will need to test the existent support and adapt your expectations.

You can submit text sentences to the Microsoft Translator API and then pipe the results to the Microsoft LUIS and then trigger your actions.

The above workflow/solution works with text.

Developing a plugin does not solve all the tests and above issues and if you need (want) to interface with ARC there are other available methods: Http Server, Read & Write from a file on the disk (EZ-Script), Javascript plugin.

#2   — Edited

Sorry, I may have mistaken for you but I'm from Japan and I'm working in Vietnam because my father is Vietnamese. Thanks again for the answer, I will try it and if there is any question, please help me

#3  

I also emphasized that I have the espeak-vi package download and use Azure speech service and basically still speak even though it is still buggy. And I think it's difficult for me when I don't know the specific to work with Vietnamese but if you think that is an unnecessary question and think it easy, please help me. I just don't know, I need to learn. It's sad when you say that!!!!!!

PRO
USA
#4   — Edited

Asfaik espeak is a Speech synthesis i.e. converts text to Speech. Speech Recognition is the opposite converts Speech to Text.

If you are looking Vietnamese Text to speech  windows 10 supports (similar answer to your question regarding Korean support): User-inserted image

You can install the Vietnamese language support, and use ARC Speech synthesis control, and i believe it works much better than the espeak package.

As i mentioned before (Korean post) windows support (above add-on) can't be used for Speech to text.

So are you trying to find a solution for Speech to text or Text to speech ?