Sensor-I

98% of crashes in the U.S. occur as a result of user error. Sensor-I is a facial and object detection system that strives for zero user error crashes per year.

Comment

Inspiration

We live in a big college town, and parties and underage drinking are very common. Our biggest fear would be losing one of our peers to the millions of accidents that happen every year. Car accidents occur every day in San Antonio. Preventing accidents and saving lives is what inspired us to take on this challenge.

How we built it

We used ML frameworks like opencv and Pytorch for computer vision. We also used packages like numpy and yolov5 for scientific computation and the detection of objects.

Challenges we ran into

1.Troubleshooting our laptops 2.Interfaces that did not sync with current versions data languages 3.M1-Chip and Intel incompatibilities with packages 4.Getting over the first-time learners hump

Accomplishments that we're proud of

  1. We stuck together as a team to help each other troubleshoot, learn new skills, and organize our data
  2. We were encouraging, took breaks when needed, and played to our strengths
  3. We all became better software developers as a result of our troubleshooting

What we learned

  1. How to install packages and interfaces through the terminal
  2. How to push and pull on GitHub
  3. How to work with different API's
  4. How to troubleshoot ## What's next for Sensor-I
  5. Using stronger ML models with stronger algorithms for more accurate real-time detection
  6. Upgrade our hardware sensors
  7. Use government databases to program Sensor-I with software to detect licensed drivers in said states
Share this project:

Updates