Check this out: Your doorbell (if you use a Chui unit) can recognize you, your family, and your friends. Also your mail carrier, your ex, and those thoughtful but persistent Jehovah’s Witnesses.
Chui is a facial-recognition doorbell and security system. It’s being crowdfunded and currently has 110 percent of its raise locked down.
Here’s what it does:
- Unlock your door for the right people
- Act as an audio/video communication channel for your front door
- Give pre-recorded messages to specific people
- Act as a “do not disturb” sign
- Detect motion and record security events
- Integrate with other “smart” home devices
And of course, there’s more. But why read about it when you can see this nifty demo video?
Naturally, we had a few questions about this gadget, so we got in touch with creator Shaun Moore via email last week. Here’s how that went down:
VentureBeat: Obviously, facial recognition raises questions about personal privacy. Where does Chui store data on faces, and what’s the device-to-device communication like?
Shaun Moore: There is no data stored on the device, which mitigates any risk associated with a stolen Chui. The data is fully encrypted and sent to be processed and stored in the cloud. This will prevent anyone from accessing any user data via the device itself. Furthermore we are working with a group of cyber security specialists to utilize their government grade end to end encryption in our data storage and transfer protocols.
VentureBeat: How is this system superior to other keyless entry systems such as Lockitron?
Moore: It opens up different avenues for users to utilize our technology. You no longer need to require your users/customers/members to have a certain kind of smartphone, to download a mobile application, or carry something with them at all time. Eliminating the obligation to carry around loss prone and transferrable entry products – fobs, cards or phones to grant access. You can unlock a door or enter a gym simply by being in front of Chui.
VentureBeat: How does the facial recognition technology match up with similar tech from, say, Apple or Facebook or Google?
Moore: Our technology is 99.6% accuracy in a controlled environment. We have invented several proprietary solutions to augment the accuracy and we’ve added second layers of security for a users peace of mind.
VentureBeat: Is the system hackable? If I had a picture of you or a mask that looked like you, could I get into your home?
Moore: Currently we utilize innovative computer vision techniques to ensure that the system will not be fooled by a picture or video. For the users who are more involved with security, we offer second layers of verification for extra secure spaces. Furthermore, some of the techniques we’ve implemented, such as a liveness check, measures micro movements on the person’s face which serves to prevent hacking via masks. We’re continuing to test techniques such as obtaining the face’s heat signature to prevent other spoofing attempts.
VentureBeat: What kinds of people are going to be your earliest adopters?
Moore: We have received orders from 13 countries, and our buyers have ranged from young tech savvy adopters to concerned parents and people desiring to use it in commercial environments.
VentureBeat: Are y’all going to be raising financing aside from crowdfunding anytime soon?
Moore: We are planning to initiate our seed round in the next few months.