0

My team is buildung an app which is using the SpeechRecognizer for Android. The app is acting as a way to talk to the user's banking account, means the user is able to ask the app "What is my balance?". Then the app will get the data from a banking backend (speech is converted to a text intend) and present the data to the user via speech and in a chat bot. We are unsing the SpeechRecognizer component of Android for handling transforming speech to text and the way back.

Another command for the app could be: "Transfer 50 Euro to the account of my wife with the number 12314567893231."

My questions are:

  • Does the processing from Speech-to-Text and from Text-to-Speech occur on the device or on the Google server?
  • If processing is done on the server: what data is cached / stored on the Google server in this case?
  • Is the data stored in such a way that conclusions about account balances and receivers are not possible?

Any ideas about this topic?

Roland Oehen-Kanzow
  • 1

1 Answers1

0

Does the processing from Speech-to-Text and from Text-to-Speech occur on the device or on the Google server?

Documentation says:

The implementation of this API is likely to stream audio to remote servers to perform speech recognition. As such this API is not intended to be used for continuous recognition, which would consume a significant amount of battery and bandwidth.

On the other questions, you'd better contact Google support, they do not give answers there. And will not likely answer you. If you really care about privacy you'd better use your own implementation.

Also discussed at Offline Speech Recognition In Android (JellyBean)

Nikolay Shmyrev
  • 24,897
  • 5
  • 43
  • 87