This is one of my first times creating an ML algorithm completely from scratch and utilizing it in a project!
ASLVision uses OpenCV and a Random Forest Classifer (self-made) to help and test users' ASL alphabet signing.
Access the deployed version here! https://aslvision-frontend.onrender.com/
Here is a quick demo video:
ASLVision.Video.Demo.mp4
🔤 Real-time ASL alphabet recognition using webcam
🧠 Custom-trained Random Forest classifier
🧪 Test mode with scoring and timers
🎥 Frontend webcam integration (MediaPipe Hands)
Frontend: React + TailwindCSS
Backend: Python + FastAPI
Web Camera Operations: OpenCV, MediaPipe Hands
ML Training: Pandas, Numpy
This is Phase One of this project! The next steps of the project include
- Exploring more ML algorithms that may outperform Random Forests
- Hosting Database for user's scores and preferences
- Adding ability to detecet words instead of just alphabets