VocalZoom’s Microphone ‘Reads Lips’ To Improve Sound Quality

Our voice is a fundamental engine of communication and has the potential to be the most pervasive Natural User Interface (NUI). Yet one of the main inhibitors that limit the potential of voice as the choice NUI (most of us play around with Apple’s Siri for a couple of days and then revert to typing our requests) is the often poor quality of voice input, mostly due to surrounding noise.

In fact, almost all voice recognition applications (like Siri) are difficult to use in open spaces and voice communication applications (like conference calls) encounter the same problem.

Israeli startup VocalZoom has developed an optoelectronic microphone that is able to extract a person’s voice from almost all surrounding noise – including other people’s voices.

Related articles

• PrimeSense Wants To Put Tiny Affordable Motion Sensors In Everything
• Israeli Technology For Peace (And Quiet)

VocalZoom began its quest to create their microphone in 2007, when CEO Tal Bakish left his day job at Cisco Systems to step into the startup world. In November 2009, the company secured its first private investor. Since then VocalZoom has been busy at work to finalize its prototype for the optoelectric microphone, securing funding along the way from big names like 3M.

A microphone that can read lips

VocalZoom’s patent-pending technology enhances speech by combining two independent sensors: A standard acoustic microphone and an optical sensor, capable of detecting voice from a specific user, unaffected by surrounding acoustic noise.

Helping Siri understand you

The optical microphone uses a senor to detect sound vibrations on the face of the person talking, and then uses the frequencies of the speaker’s voice to “visualize” what the person is saying. It’s almost like reading the lips of the person you are talking to, without actually hearing what they have to say. The optical microphone feedback is then subtracted from the noise that the regular acoustic microphone picks up, so that the speaker’s voice is isolated from any background noise.

Bakish tells NoCamels: “If you put your hand on your throat, you can feel vibrations when you talk that radiate all over your face. Although you may not realize it, your voice vibrates the skin on your face, while external noises are unable to have the same effect.”

“In a nutshell,” says Bakish, “VocalZoom’s technology creates a ‘virtual cube’ in space, sensing sound from only within the cube,”

He adds: “This results in a significant speech enhancement and precise speaker isolation, which are the key element missing today to enable mass-usage of voice driven applications.”

Microphones are everywhere

Microphones are present in most of the technologies we use every day and these are increasingly driven by voice activated systems, which is why VocalZoom believes its chances are especially promising. According to Bakish, one major automobile company has already invested in VocalZoom’s product to improve their Bluetooth carkits. “And there has been interest from major OEMs (original equipment manufacturers), companies that provide parts for communications devices and portable computers,” Bakish says.

Bakish believes that VocalZoom has been able to set itself apart from other clear-voice technologies for two main reasons. The first is that, unlike other attempts to improve voice recognition that only try to understand the noise and then adjust the technology, the optoelectric microphone focuses its efforts solely on the speaker. Second, unlike other noise-reducing technologies, such as bone conduction headphones, VocalZoom’s microphone is a remote device that can be installed in a number of different places like telephones, cars or computers. Bakish believes that the optoelectric microphone is what programs like Siri and Google Voice need to function properly in all kinds of noisy environments, and even to understand different kinds of accents.

Currently, optical microphone technology has been primarily applied in Optoacoustic’s technology for use in medical, industrial and environmental fields. Yet, as Bakish points out, VocalZoom is the first to try out the micrphone in the world of voice recognition for computerized devices. The optical microphone was picked up additionally by the sound-giant Sennheiser for similar applications, but without the versatility and compactness of VocalZoom’s prototype.

Hearing clearly in the future

Currently, VocalZoom’s microphone is still in the prototype stage, with the company working hard to shrink the microphone and to make the technology affordable.

Along with funding from a major automobile company in the United States and 3M, the VocalZoom project also received a grant from the Chief Scientist of Israel’s technology incubator of about $600,000, 15 percent of which is from private investment. Bakish believes that his company has enough monetary support to finish the prototype goal set before them, and hopes to earn more funds through sales once the product is launched.

Other than the field of optoelectric microphones, Bakish has also been part of several other startups, such as AdDonate, Anagog and Biocatch.

Photos: timparkinson (main) | vasile23 (Siri)