Optical human-to-machine communication sensors promise 'near-perfect' speech recognition: Page 2 of 2

October 25, 2016 //By Julien Happich
Optical human-to-machine communication sensors promise 'near-perfect' speech recognition
Headquartered in Israel with a sales office in California, startup VocalZoom was founded in 2010 to focus on Human-to-Machine Communication speech enhancement challenges. It started product development in 2013 shortly after it had signed strategic agreements a large automotive OEM and with 3M Corporation, sampling its first engineering samples of its VocalZoom HMC sensor last summer.

"For biometrics, all the processing can be done locally on the VocalZoom chip, by not having to transit to the device's main processor, data is much more secure", emphasizes Bahalul.

By detecting the direction of arrival, the sensor can also verify that the person of focus is in the right direction and at the right distance, say in ATM or automotive use cases, to only listen to a particular user.  

VocalZoom has recently completed a design-in phase with speech recognition provider iFLYTEK who plans to launch a headset based on its technology before the end of the year. Even in a busy call centre, the audio fusion performed in the headsets would pick up a clean signal from each individual call receptionist. The two companies are also collaborating on the development of an automotive voice control and voice biometrics product, based on a far-range version of VocalZoom’s HMC sensor.

Bahalul told us the company was expecting its sensor to be integrated into motorbike helmet headsets by the end of 2017 and that it could be prototyped into a car mirror or into a car infotainment's system by the first half of 2017.

For now VocalZoom is open to different business models.

A standalone HMC sensor unit.

"We could sell a complete sensor module with the laser and the ASIC on a small PCB, or for high volume OEMs, we could license the technology so they could manufacture everything and optimise the sensor, for example to integrate it into smartphone cameras" said Bahalul.

"We can perform all the audio processing on our ASIC, so MEMS microphones could share the ASIC. We are looking at multiple vendors to shrink the technology to their standards. When you actually look at the cost of the individual sensors that our solution could replace in a smartphone, we could offer a much better cost structure" he concluded.

Visit VocalZoom at www.VocalZoom.com

Related articles:
Optical mic 'hears' wider sound spectrum

Vous êtes certain ?

Si vous désactivez les cookies, vous ne pouvez plus naviguer sur le site.

Vous allez être rediriger vers Google.