Inspiration
Introducing NokNok, the smart door lock that is controlled by the most unique password of all - the human face. NokNok is a seamless, easy to use interface that takes home security to a whole new level. By utilizing facial recognition technology, NokNok can determine if you are an authorized user, and allow you to enter your house - no need to fumble around with passcodes or remote controls.
How it works
NokNok users simply go up to the front door and ring the doorbell, just as they normally would on any other door. NokNok's 5mp camera will then take a picture of the user and run it through a facial recognition software (API provided by Face++, via Mashape). If the user matches one in the "approved" database, the door sends a command to the SmartThings hub, instructing the included smart lock to activate.
Challenges I ran into
One of the biggest challenges we faced was learning how to utilize the face recognition data from the Lambda Labs API and turn it into useful data for our software. The API outputs the data as JSON, a language we were initially unfamiliar with. In addition, we initially faced difficulty sending HTTP get requests using Unirest, an HTTP request client library.
Accomplishments that I'm proud of
We are very proud of our app's combination of multiple APIs to provide a cohesive solution to an everyday problem. In addition, we are proud that we were able to utilize facial recognition technology, something that several years ago was only possible by governments and large corporations, to aid in our solution.
What I learned
Through this hack, we gained a great deal of knowledge in how to utilize and parse JSON in order to obtain useful data for our app. One of our team members previously had no experience with Python, so this hack required him to pick up a new language in order to write programs for the Raspberry Pi. In addition, we learned a lot about how to insert APIs into our program, something we got to experience for the first time. Lastly, we learned about the mechanics of facial recognition technology, and how our API processes faces in order to provide a match.
What's next for NokNok
Our long-term goals are to improve the front-end user experience and providing more intuitive control to users via a mobile app. Through the app, users can be notified who is at the door, and they can choose whether or not they would like to answer it. They can also set up a vacation mode to remotely see who is ringing their doorbell, and choose to remotely grant visitors access if they so wish. In addition, we would like to integrate with Facebook in order to utilize users' friends lists and recognize their friends' faces based on Facebook photos and profile pictures. This way, you can always know who is at the door, and can make a smart decision whether to answer or not.
Log in or sign up for Devpost to join the conversation.