Peter Yungman

🚀 Projects

• Ultimate (Frisbee Angle Analyzer)

  Mobile app that analyzes throwing mechanics in ultimate frisbee using computer vision. Developed an app that uses the phone camera to track wrist angles and throwing form. Provides real-time feedback on separation degrees, reachback, elbow timing, and wrist speed. Designed with the goal of helping players improve consistency and prevent injuries.

  • Python
  • MediaPipe
  • OpenCV
  • Flask
  • React Native
  • Firebase

  
• Signly

  AI-powered sign language interpretation tool bridging accessibility gaps. Designed to interpret ASL into text in real-time, helping Deaf and hearing individuals communicate more easily. Focused on computer vision recognition of handshapes, movements, and context. Prototype stage involved model experimentation and dataset exploration.

  • Python
  • TensorFlow
  • PyTorch
  • MediaPipe Hands
  • Natural Language Processing

  
• WingingFlights

  Cheap flight alert service with daily scraping and personalized notifications. Built a system where users sign up with email, travel dates, duration, and price range. Each morning, the app scrapes flight data and emails users any matching low-cost options. Automates planning and reduces hours of manual searching for deals.

  • Python
  • BeautifulSoup
  • Selenium
  • SMTP
  • SendGrid
  • SQLite

  
• EchoGuard

  Wearable night-time obstacle alert to help seniors navigate safely in the dark. Team-built Arduino device that mounts on a hat kept by the nightstand. When the wearer gets up, an ultrasonic sensor detects nearby obstacles and triggers a piezo buzzer and light to warn of collisions—especially useful in low light. Includes tunable distance thresholds and alert patterns to reduce false alarms.

  • C/C++
  • Arduino