Tuesday, October 6, 2015
Linux Speech Recognition Using Google Speech API: Google2Ubuntu
Google2Ubuntu is a tool that lets you control your computer using voice commands via the Google speech recognition API.
The project is not new but it wasn't updated for about two years, until recently, when its developer added some new features along with English support (French was already supported).
Update: the latest Google2Ubuntu adds German, Spanish, Traditional Chinese, Portuguese (BR and PT) and Italian support. The new version also includes a menu to change the language, a dictation mode, new setup window along with some other improvements.
Update: the latest Google2Ubuntu adds German, Spanish, Traditional Chinese, Portuguese (BR and PT) and Italian support. The new version also includes a menu to change the language, a dictation mode, new setup window along with some other improvements.
Currently, Google2Ubuntu comes with the following features:
- available in English, French, German, Spanish and Italian for now;
- internal commands: can tell you the current time, battery level or read the text selected by the cursor;
- external commands: window actions (close, hide, switch), open various applications (web browser, terminal, file manager, open specific folders such as Music, Pictures or Documents), cut, copy, paste;
- 4 external modules are available, which allow you to search Google, Wikipedia, or Youtube, get the current weather or find out the meaning of a word;
- dictation mode;
- the commands are customizable so you can change them or add your own custom commands.
It's worth mentioning that since Google2Ubuntu uses the Google speech recognition API, it needs a working Internet connection.
Unfortunately the Google voice recognition API doesn't like accents so if you don't speak English or French natively, this tool will fail at times. Or at least, that was the case in my test and that's why I didn't record a video. But you can find a Google2Ubuntu video demo HERE - the video is in French but it should be enough to get an idea on what this application can do.
Unfortunately the Google voice recognition API doesn't like accents so if you don't speak English or French natively, this tool will fail at times. Or at least, that was the case in my test and that's why I didn't record a video. But you can find a Google2Ubuntu video demo HERE - the video is in French but it should be enough to get an idea on what this application can do.
Install Google2Ubuntu
Google2Ubuntu is available in a PPA for all supported Ubuntu versions. Add the PPA and install it using the following commands:
sudo add-apt-repository ppa:benoitfra/google2ubuntu
sudo apt-get update
sudo apt-get install google2ubuntu
I tested the application under Ubuntu 13.10 and 14.04 so I'm not sure if it works properly with older Ubuntu versions.
How to configure and use Google2Ubuntu
1. Once installed, you need to set up a keyboard shortcut for triggering Google2Ubuntu. When you use this keyboard shortcut, the Google2Ubuntu speech recognition will be activated, listening for your command (a sound and a notification will be displayed, telling you when to speak).
Let's add the keyboard shortcut:
- in Unity/GNOME, open System Settings > Keyboard > Shortcuts, then click on Custom Shortcuts on the left, click the "+" button and add "/usr/share/google2ubuntu/google2ubuntu.py" for the command and "Google2Ubuntu" for the shortcut name. Then assign it a keyboard shortcut (I've used F7 in my test).
- in Xfce: open Settings > Keyboard and on the Application Shortcuts tab click "Add", then add "/usr/share/google2ubuntu/google2ubuntu.py" for the command and then enter a keyboard shortcut when prompted:
Make sure the keyboard shortcut you've entered above isn't already in use!
2. Next, you can take a look at the available commands, edit them and/or add your own by opening the Google2Ubuntu-Manager from the menu / Dash:
You can find out more about what commands you can use with Google2Ubuntu by visiting its GitHub page.
3. Now when you want to trigger the Google2Ubuntu speech recognition system, press the keyboard shortcut you've set up under step 1, then speak the desired command.
When pressing the keyboard shortcut, a sound as well as a notification will let you know when you can speak:
Once the command is completed, a notification is displayed:
That's it.
If you want to help with the Google2Ubuntu development, report bugs or get the source code, see its GitHub page.
via ubuntuforums
Labels:
google,
linux,
speech recognition,
Ubuntu
Subscribe to:
Post Comments (Atom)
The ppa for this application I received a 404 error. not there.
ReplyDelete