WeChat public platform Message Interface Development (40) speech recognition

Public platform development mode speech recognition public platform Message Interface speech recognition

Public platform development model Speech recognition Public PlatformMessage Interface speech recognition

The speech recognition process on the public platform is quite complex.

1. obtain voice file data

Currently, you can obtain voice files through simulated logon.

II. speech encoding

Generally, the interface requires a specific encoding, so transcoding is required.

3. submit data to a third-party interface

Obtains the recognition result and confidence level, and obtains the highest value.

 "; Foreach ($ data ['hybrid'] as $ I) echo"
". $ I ['utterance ']."
"; Echo"";} Else {echo"Identify errors";}?>

IV. semantic recognition

You can use word segmentation to analyze the results through syntax rules.

V,

Others: Text to Speech

Google Text to Speech API

Base URL: http://translate.google.com/translate_tts
It converts written words into audio. It acceptsGETrequests.

GET

Q
The query string to convert to audio

Tl
Translation language, for example, arfor Arabic, oren-usfor English

Ie
Encoding format, use defaultUTF-8

ExamplesThis is an example for Arabic "?????? ????? "

http://translate.google.com/translate_tts?ie=UTF-8&q=%D8%A7%D9%84%D8%B3%D9%84%D8%A7%D9%85%20%D8%B9%D9%84%D9%8A%D9%83%D9%85&tl=ar

This is an example for English "Hello World"

http://translate.google.com/translate_tts?ie=UTF-8&q=Hello%20World&tl=en-us

Public Platform Navigation