If you’re following Alexa Voice Service news, you’ll see that the number of options that are available for hands free / far field solutions is increasing. Currently, there are seven kits advertised on the AVS Developer site. On those, there are at least four different wake word engines (Amazon’s own, Intel’s, Sensory’s, and Kitt.AI’s).
This is quite an advanced compare to four years when the readily available DSP chip for far field and acoustic echo cancellation was not available. Sure, there were lots of options on the market but none that could be deployed on a chip without considerable expense and effort.
Likewise, with wake word, there were only a few players and some of them only worked for embedded solutions, rather than on target. Now, we have quite a few options available, especially for host processor based solutions.
What’s great about having so many different technologies coming out is that it is going to drive down the cost for hardware companies to add far field voice. The danger to Amazon is that this is also seeding the field for embedded Google Assistant, Bixby, or other technologies. Amazon will need to continue to move as quickly as possible with features to ensure that others won’t catch up.