On 10/03/2017 18:33, John Rumm wrote:
On 10/03/2017 13:55, Dave Plowman (News) wrote:
In article ,
Adam Funk wrote:
If the TV can be voice controlled, it must be able to separate speech
in the room from stuff coming out of the TV itself, right?

Think there's a difference from recognising a clear command to
understanding all speech.


Unlike for voice command processing, the TV does not need to understand
the "bugged" speech - only upload it.



The TVs tend not to understand the speech and do the same as apple and
M$ ie. the upload it to a server for recognition.

So if you want to listen in you can hack the server as long as the user
is using the voice recognition.
If they have disabled voice recognition in the menus then you need to
hack the TV.