Inspiration

As Computer Science majors, we constantly struggle with keeping track of our posture and not straining our eyes. We wanted to develop an application to help with better human-computer interaction and maintaining a healthy lifestyle

What it does

ErgoBlink detects bad posture, blinking patterns, and reminds you to take breaks for healthy screen time habits.

How we built it

ErgoBlink was built using computer vision, machine learning algorithms, and deep learning techniques. As well as OpenCV and MediaPipes for the face recognition, we trained our own model for the posture detection and used EAR algorithm for blink detection.

Challenges we ran into

Developing accurate posture & blinking detection algorithms, & integrating them into a seamless user experience as well as package version control.

Accomplishments that we're proud of

Successfully creating an app that promotes healthy habits & reduces negative effects of prolonged screen time. As well as a corresponding front end.

What we learned

Deepened understanding of computer vision & machine learning, & importance of prioritizing user health in tech.

What's next for ErgoBlink

Expanding platform to incorporate additional features & improving accuracy of posture & blinking detection algorithms.

Built With

  • elvenlabs
  • jupyter-notebooks
  • mediapipes
  • nextjs
  • openai
  • opencv
  • react
  • scikitlearn
  • tailwind.css
Share this project:

Updates