Re:Facebook is able to identify music and TV when writing st...
by remon on Monday November 02, 2015 @11:31AM
(#10492817)
> because nobody is doing this kind of processing client-side.
This is a half truth at best. Audio recognition can be based on one of two techniques. The first is audio watermarking which is completely client side (albeit not applicable for this FB feature) or fingerprinting. They'll likely use the latter which records a segment of audio, does the rough audio equivalent of a hash on it and then send it the server for an optimised pattern match. It's somewhat unlikely they'll send raw audio to their servers although that's an assumption.
Re:Facebook is able to identify music and TV when writing st...
by Gnewt on Monday November 02, 2015 @12:57PM
(#10493638)
I mean, it's an assumption that's got some known counterexamples. For example, Google Now stores the raw audio from every time you activate it / dictate a question on an Android device. It's not out of the question to expect that Facebook might be keeping raw audio data to process later / test.
...by recording the conversation so it can be uploaded to the server that does the recognition, because nobody is doing this kind of processing client-side.
Just like the idea that "streaming" is somehow different than "downloading", the commonly-repeated claims that this kind of technology is only "processing" without mentioning where the data is processed is a carefully constructed lie designed to frame the discussion.
The use of propaganda to confuse how their product works suggests that Facebook (et al) knows there would be resistance if they offered the same features plainly. In my opinion this should count as mens rea.