CamSafe

• 

  facial recognition code

• 

  facial recognition training script

Inspiration

People want their families to be safe and their home to be secure. We developed a facial recognition system to recognize unknown and known faces to allow easy access to see who is behind your door.

What it does

Imagine an intruder trying to break into your house. They yell through your door to open the door, pretending to be a family member. You hear the voice and open the door without ensuring who it is outside the door. The door is opened and the intruder is let in, leading to an unpleasant ending. What could have been done to not let this scenario or any other break-in occur? Our product CamSafe, provides the essential solution. CamSafe is a low-cost home security system that uses facial recognition technology, outputting a live video stream and sound output of the person outside your door. The facial recognition images the face of the person, and outputs if the person is known in the system or not. In the case of it being a known person, will result in an audio output to your device or speaker stating who is outside your door. But as seen in the first scenario, if it is an intruder then the output will be an unknown person informing you not to open the door. This will keep security management and let you rest easy knowing your family is safe.

How we built it

We used Python to develop a facial recogniztion system that scans an individuals face and relays that output in the form of a name on the live imaging video

Challenges

Maching Learning Training Set not accurate Real Time Streaming Video

Accomplishments we are proud of

Acheiving a facial recognition software complete with machine learning/video output/audio output within 10.5 hours

What we learned

How to effectively use machine learning to solve problems

What's next for CamSafe Security System

We are always thinking about future improvements to our system. Our vision for the future is to add integration with other softwares to better communicate the audio output of the facial recognition system, including Google Home and Alexa. We will also make our system interaction more informative to allow more information about the person rather than the status of known or unknown. Lastly, we will keep pushing further developments to make our facial recognition software more accurate, because safety is the number one priority.

Built With

• hardware
• machine-learning
• python