Is there any speech in the text API or some kind of “hack” with which I can use Google speech for text tools?

I think Google’s speech for text tools (automatic transcription of Google Voice voicemail, automatic video signing on YouTube, etc.) is impressive.

I really looked to see if Google provided access to it through the API, and it seems not (I don't blame them!). The cloud computing service providing a speech function for text will be pretty cool.

Is there some kind of “hack” that I can use to access speech in text. My architecture basically comes down to this - a short 15-20 seconds wav / mp3 / other clip as input, the output is plain text.

Any ideas from people?

+4
source share
6 answers

There are many text based APIs. Just because Google doesn't make them accessible doesn't mean you're out of luck.

Here is a good example for C #. You can search for others for your platform if it is not .NET.

http://cmusphinx.sourceforge.net/

+3
source

Check it out: http://mikepultz.com/2011/03/accessing-google-speech-api-chrome-11/

I am currently trying to implement an API in PHP.

- Seth

+2
source

It is available in HTML5 via Chrome 8 or Opera: https://docs.google.com/View?id=dcfg79pz_5dhnp23f5&pli=1

Google voice technology is also available through the Android API on your Android phone.

Other products, such as Sphinx, are speech recognition engines that work best in specific areas, rather than “unconditional” speech text.

+1
source
+1
source

You can take a look at the following implementation using C # - I used Mike Pultz's link.

https://github.com/seigneur/Voice-Biometrics I used Sox to convert to flac, created a small SOX script to break it into pieces.

+1
source

If you really need google output ... Here is the Hack method

Have you thought about creating a messaging engine? Essentially, it calls your voicemail google ... plays mp3.

Run the output through https://code.google.com/p/google-voice-java/

The best answers.

+1
source

Source: https://habr.com/ru/post/1309489/


All Articles