5 ms·

I find it odd that Google is going to take so long to get this out the door - "later this year" seems like ages. Did they start on the hardware that late? Amaz

by pbnjay 10y ago

I find it odd that Google is going to take so long to get this out the door - "later this year" seems like ages. Did they start on the hardware that late? Amazon has what, 6 months to get more competitive on the search/trivia front? or this is going to kill it.

martincmartin 10y ago
The hardware for this is the easiest part. Understanding voice that's far from the microphone (sound gets muffled, echos off walls, etc.), with other sounds in the background, is significantly harder than understanding speech near the microphone.
- harigov 10y ago
  Didn't they already solved that problem for implementing "OK Google"?
- pbnjay 10y ago
  "OK Google", Google Now, their voice recognition tech, and NLP recognition are far far past Echo's version. So if the hardware is that easy this should have been out yesterday. The even mentioned earlier in the keynote how they've been training it Google assistant with background noise.
  - hiddencost 10y ago
    Speech recognition is a lot harder than you'd think, and the problems of far field speech recognition especially. All of google's in-house ASR (and wakeword) has been dedicated to near-field devices. Speech is not a "solve it once and it works everywhere" kind of problem. I would ballpark that Google needed to collect 10k hours of _far-field_ speech to train this to a level that was acceptable for the enormous variety of noise conditions and accents that a far-field system needs to work on. That scale of data collection takes time and a lot of trench-warfare effort.
    - pbnjay 10y ago
      Sure I understand, but my point is that they should be ahead of the game on basically everything else. Echo was released over a year ago so surely Google has been working on far-field tech since _at least_ then. To still be aiming for end of year seems like some serious missteps and catch-up going on.

by pbnjay 10y ago

martincmartin 10y ago

The hardware for this is the easiest part. Understanding voice that's far from the microphone (sound gets muffled, echos off walls, etc.), with other sounds in the background, is significantly harder than understanding speech near the microphone.

harigov 10y ago
Didn't they already solved that problem for implementing "OK Google"?
pbnjay 10y ago
"OK Google", Google Now, their voice recognition tech, and NLP recognition are far far past Echo's version. So if the hardware is that easy this should have been out yesterday. The even mentioned earlier in the keynote how they've been training it Google assistant with background noise.
- hiddencost 10y ago
  Speech recognition is a lot harder than you'd think, and the problems of far field speech recognition especially. All of google's in-house ASR (and wakeword) has been dedicated to near-field devices. Speech is not a "solve it once and it works everywhere" kind of problem. I would ballpark that Google needed to collect 10k hours of _far-field_ speech to train this to a level that was acceptable for the enormous variety of noise conditions and accents that a far-field system needs to work on. That scale of data collection takes time and a lot of trench-warfare effort.
  - pbnjay 10y ago
    Sure I understand, but my point is that they should be ahead of the game on basically everything else. Echo was released over a year ago so surely Google has been working on far-field tech since _at least_ then. To still be aiming for end of year seems like some serious missteps and catch-up going on.