2 ms·
> 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
by remon 11y ago
> 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.
- Gnewt 11y agoI 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.