Waeky-Waeky
This project was initiated to solve specific technical and domain challenges. Processing webcam frames in real‑time with an AI model without lagging the computer or draining the battery excessively.
About This Project
Processing webcam frames in real‑time with an AI model without lagging the computer or draining the battery excessively.
Optimized frame extraction, utilised lightweight eye‑aspect‑ratio computation via OpenCV in a background process and implemented a Tauri/SolidJS front‑end with customizable alerts.
Created a working prototype for driver safety that actively alerts users upon micro‑sleeps.
Machine Learning Engineer
2024
Public
Personal
Technology Stack
Project Story
Processing webcam frames in real‑time with an AI model without lagging the computer or draining the battery excessively.
Optimized frame extraction, utilised lightweight eye‑aspect‑ratio computation via OpenCV in a background process and implemented a Tauri/SolidJS front‑end with customizable alerts.
Created a working prototype for driver safety that actively alerts users upon micro‑sleeps.
Insights & Takeaways
Highlights
- Case study content natively baked into the project dataset.
- Clear storytelling built around the specific problems faced and the technologies used.
Challenges
- Strict focus on performance and maintainability.
- Selecting standard tools to ensure scalability: Python, OpenCV, Flask, Tauri, React, Rust, Tailwind CSS
- Optimized frame extraction, utilised lightweight eye‑aspect‑ratio computation via OpenCV in a background process and implemented a Tauri/SolidJS front‑end with customizable alerts.
Lessons Learned
- Created a working prototype for driver safety that actively alerts users upon micro‑sleeps.
