Skip to content

PasaOpasen/SpeechLogger

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

51 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

SpeechLogger

The simple application for logging (and translating) all speech from your microphone and speaker.

I think it would be very useful when you communicate with somebody foreign person by Discord or else app. It can help you to have better understanding your interlocutor if you know his language like I do english)

First result

Scope:

  • Translation messages from any languages supported by Google into certain choosen languages at once

  • Recognition and translation the speech on choosen language from microphone

TODO:

  • recognize all speech from microphone automatically (not only on demand)

  • recognize speech from speakers (it will be hard to do)

  • write program into desktop app

  • add text logging in file

1

How to run

You can download this app as compiled (see releases). Unpack and run .exe file.

Notes

  • Not correct showing of arabic words (reversed and upper-cased) exists because of bags with windows terminal. But if u will copy this text and paste it into text editor / google / messanger, it would be normal.

Second result

What's new:

  • recognition speech from speakers too (if u have a speaker supported callbacks)

see

How to run

I cannot compile it now because of problems with soundcard, but u can run it by Python:

  1. Download Python from official site (choose correct operation system)

  2. Open cmd.exe

  3. Install necessary packages via commands:

pip install numpy
pip install scipy
pip install textblob
pip install pyaudio
pip install soundcard==0.3.3
pip install speechrecognition
pip install termcolor
pip install colorama
  1. download and run this file

Notes

  • if u have some problems with installing pyaudio, try to get solutions here

Third result

What's new:

  • json files with basic settings and supported languages
  • trancription (grapheme to phoneme) by epitran

TODO firstly:

  • transcript persian words better

im

How to use

  1. also install epitran by command (cmd.exe):
pip install epitran
  1. download release

  2. unpuck and run .py file

4th

What's new:

  • TextBlob API was replaced with better googletrans API
  • English is transliterated by pysle, farsi --- by PersianG2p

1

How to use

pip install googletrans
pip install pysle
pip install PersianG2p

Download release, unpuck and run .py file.

News

Since 2 June 2020 I'm working with this Telegram translator bot.