Inspiration
- Unfair racial bias towards people
- People being treated unfairly (especially at security)
- Wanted to create an application that not only tackles this but still helps people feel safe
## What it does
- Input: Live feed from the camera
- Output: Classification of whether the person is 'valid' in the database or not + led light flashes red or green
## How we built it
- Website: tensorflow.js, javascript, html, css
- Computer Vision / Machine Learning: Python, opencv, tensorflow
## Challenges we ran into
- Can't get the application to be hosted on IBM Cloud (Even people from IBM couldn't help us either)
- Connecting output of Computer vision part and website part (resulted in incompatible photos)
## Accomplishments that we're proud of
- Got video feed working/displaying via laptop webcam and webcam connected to pi
- Made our own dataset
- LED flashing after the result of recognizing faces
- Recognize faces (draw a bounding box around it)
## What we learned
- tensorflow
- ibm cloud
- opencv
- raspberry pi
- IOT
- First hackathon for some members
## What's next for SafetyNet
- Finish connecting computer vision and the website component
- Send email / notify Alexa or google home after classification
- Improve UI
Log in or sign up for Devpost to join the conversation.