INTERVIEW: Todd Mozer, CEO, Sensory

INTERVIEW: Todd Mozer, CEO, SensoryMobile ID World President Peter O’Neill recently spoke with Todd Mozer, CEO, Sensory. The conversation begins with talk about the latest versions of Sensory’s two flagship technologies, TrulySecure 3.0 and TrulyHandsfree 5.0, which offer contactless authentication and user interface respectively. The conversation turns to how the face and voice biometrics offered by Sensory are well positioned to take advantage of the booming Internet of Things space, the company’s collaboration with Amazon on its Alexa AI technology, the secrets to good Wake Word technology, and the factors behind Sensory’s growing revenue, which is set to double this year, according to Mozer.

Read the full interview with Todd Mozer, CEO, Sensory:


Peter O’Neill, President, FindBiometrics: You are just launching a new version of TrulySecure 3.0, can you please tell our readers about this?

Todd Mozer, CEO, Sensory: Sure. TrulySecure is a very exciting product for us. It is a fusion of face biometrics with voice biometrics so it is a unique set of technologies that you can use either one independently or both together. The goal of TrulySecure is to be really fast, really accurate, and really convenient at doing biometrics and by combining biometrics together, we think we have achieved that.  3.0 is our third generation of the technology and we are using convolutional neural nets and other state of the art proprietary approaches. With 3.0 we’ve introduced a number of improvements in accuracy, in anti-spoofing, in liveness detection, and in being able to work better with glasses. Some approaches for face authentication might require that you remove your glasses but we have a new technique that trains with and without glasses, so you don’t need to take them on or off. There are other features like iOS improved compatibility and FIDO compliance on iOS and a few things we have done to catch iOS implementations to our Android base.

FB: What vertical markets are you aiming for with this new version of TrulySecure 3.0?

Sensory: We are seeing a lot of different opportunities in different markets. Banking and enterprise markets are certainly an area where we are getting some nice traction now. We are also seeing a lot of interest from the mobile phone community. When Samsung and Apple decide to do something pretty much the industry follows, so we are getting a lot of interest from the mobile phone manufacturers for face authentication. Our TrulySecure technologies are shipping in both mobile phones as well as being used by banks and enterprises this year.

FB: Fantastic. You also just recently launched your on-device voice interface engine called TrulyHandsfree 5.0, what were the drivers behind this launch?

Sensory: TrulyHandsfree has been Sensory’s bread and butter technology for many years. It is a very small footprint speech recognizer; a lot of companies use it for Wake Word and for small vocabulary command and control and we’ve been putting a lot of effort into this as of late. The whole IoT space has really started to boom with the success of Amazon’s Alexa and increased demand for far-field wake words that can work in a small footprint. When you are always on, always listening for Wake Word, you’ve got to be really good at rejecting everything except the right word and really responsive when you hear the right words, so Sensory has pumped increasing resources into this to stay ahead of the pack and TrulyHandsfree 5.0 is the result of this effort.

Sensory has been a neural net and machine learning company for 20+ years and as we’ve progressed we’ve learned that we need to find new ways of collecting data so that we can do deeper and more powerful nets. We’ve applied this to both TrulyHandsfree and TrulySecure.

FB: You mentioned IoT. With the tremendous focus and growth in IoT devices, what do you see down the road? Voice seems to be the technology that is almost preferred when it comes to IoT, is that your opinion as well?

Sensory: Touch has been around forever, and touch is not going to go away, but certainly voice is now really seeing a lot of expansion and deeper penetration into products. This is a natural progression because most of our communications are with voice. A lot still needs to happen; when you and I talk face to face we get a lot of cues from our vision and from our other senses and as those things start to come into play where we know who we are talking to and have more knowledge of the subject matter and the emotions and the mannerisms and the histories, the meaning recognition will get better and better and the whole experience gets better and better. We expect that will be happening in the IoT space; more intelligence from other sensors and from the history and experience.

FB: Has this been a good year for Sensory? You’ve had several big announcements and partnerships over the last twelve months. Tell us about this please.

Sensory: Yes, we have a lot of exciting things going on in an ongoing basis. From a business perspective, our revenues are going to double this year so that alone is a really nice thing. We are getting a lot of penetration in a lot of markets; we are doing extremely well in the mobile phone business, we are entering the enterprise and banking markets with some really exciting design wins and partnerships, and we are seeing success in IoT and automotive as well. Just yesterday a really interesting company called Ecobee introduced their thermostat 4.0 and the reviews are starting to come out, and they are saying it is better than Nest, and Sensory is the speech recognizer that you say “Alexa” to when you want to call up the Alexa voice system. So, we have done a lot of work in the Alexa network and with Amazon to make these kind of things happen and we are seeing a lot of opportunity in that whole space. Sensory’s speaker verification is now the best embedded verification in the world and people are starting to combine microphones and cameras together in products to further improve and create continuous ID; a lot of things that we have thought about and worked towards over the last 23 years are really starting to come to fruition so it is a fun time for us. Sensory just got acceptance of our 37th patent, so we have built up a very nice IP portfolio that targets embedded AI.

FB: Congratulations on that! Voice has become one of the dominant biometric modalities especially over that last three years even beyond IoT, what is driving all the growth and attention at this time in our industry?

Sensory: Security is obviously a really, really big deal and biometrics is much more convenient and often times more accurate approach than entering passwords. I think everybody has experienced the pain of passwords. So, voice in particular has certain advantages that are nice: you don’t need to touch devices, microphones are really cheap, it can be a far field experience, and it can be seamless in a sense in that talking without saying anything out of the ordinary can cause devices to know who you are. A good example is the fabulous Alexa network and the Echo’s and things like that where you wake them up by saying “Alexa.” Google did this recently with OK Google and embedded speaker verification into wake words is catching on fast.

Face is similar in a sense in that it can be seamless; you don’t have to do specific things for it to recognize your face, so it can be persistent and it can be seamless. And we think the more biometrics that you can layer upon each other that are persistent and are seamless the better the experience will be and the more accurate the biometrics will be because you are layering these things together.

FB: We have seen tremendous growth in voice technology and moving into areas like you mentioned Alexa, and FinTech and automotive and so many new markets. We have been following your company for many years now and it is wonderful to hear the growth that is happening, your timing seems to be perfect for your technology in the industry. So, congratulations on a wonderful year and great speaking with you today.

Sensory: Thank you very much, Peter.