3 ms·
I doubt it, that would be a good chunk of data for what I see as a fairly small feature. More likely (this is a guess, nobody outside google really knows at th
by Klathmon 9y ago
I doubt it, that would be a good chunk of data for what I see as a fairly small feature.
More likely (this is a guess, nobody outside google really knows at this point), they will use federated machine learning to figure out that something is "a song", then perhaps clean up and isolate the actual "song" part of it and send that over to a google server for processing.
But again it was just announced, so nobody really knows how this works, where the data is or goes, and what tradeoffs were made.
- angus-prune 9y agoThey announced, in the talk, that this is all done locally on the phone. The phone will have a database of, iirc, 10,000 songs locally. [edit] typos
- Klathmon 9y agoAh! Then I'm sure they have some way of making the data needed much smaller than I thought it could be!
- Anm 9y agoThey only need to store (and occasionally update) the kernel parameters for the trained deep neural net. Very small indeed.
- user-on1 9y agohow small, what technology is that?
- Anm 9y agoI don't do neural nets, but if I had to crudely estimate... 10,000 songs (the outputs) * 16 layers * 16 parameters per node * 4 for the bytes per float = 10MB + a DB of song/artist names. I'm probably underestimating the parameters per node, and overestimating the size of the layers closer to the input. Further, it's more likely structured as an LSTM than a convolutional network, since sound is a streaming source.