Let’s make Pepper and NAO speak a second language
As international talking robots, Pepper and NAO can access many languages for speech recognition, and its management is quite easy. This quick how-to explains how to set up new languages installing a language pack in a few simple steps.
Prerequisites and initial setting with Command Center
First of all, you have to be logged on Command Center, the online user interface that allows the remote management of Robots, Missions and Users.
You can log into Command Center at the following address:
Features on Command Center depend on the user profile and the type of account. Please contact the Support team for further information.
List of available languages
As a partner, you have access to all the available languages: robots with a professional account can get English (which is pre-installed) plus one additional language for free.
The list of languages is available on the “My Missions” tab on Command Center (if not please contact the Support team).
How to get new languages for Pepper QiSDK
The new language has to be available for both Pepper’s ASR and tablet display, so it is necessary to set up both Command Center and the robot tablet settings.
Get a new language on Command Center
- In the Client interface, go to “My Robots”,
- Choose a robot then click "Add missions or task" or "Add"
- Then search for the desired language and select it.
- Finally, click "Add selected missions and tasks".
/!\ If you go to “My Mission/Deploy on robots” you shall NOT get the selected language but just the latest uploaded language version regardless of the NAOqi/QiSDK version of the robot.
Install a new language on the robot via the tablet settings
- On the robot, go to the android tablet settings and swipe all the way down to “Update system/robot apps”.
- Click "Refresh": the new language should appear as an update;
- Then click on "Update all robot apps".
Local ASR: local speech recognition, only matching with QiChat pre-scripted rules, available offline.
Cloud ASR: Online speech recognition featuring free speech. Transcribes any voice input into text. Requires internet connection as it connects to third-party servers.