Sign Language Translator

Sign Language Translator is a machine learning project aims to translate hand gestures to strings. We developed this as a term project of CS 464 Machine Learning course.

How to use

1. Clone the repo: git clone https://github.com/ayberktecimer/Sign_Language_Detector.git && cd Sign_Language_Detector
2. (Optional) Create a virtual environment
3. Install dependencies: pip3 install -r requirements.txt
4. Train models: cd src && python3 Runner.py Models are saved in generatedModels directory

Dataset

https://www.kaggle.com/emresulun/sign-language-letter-images

Context Sign language letters captured by webcam and converted to grayscale images.
Letters are not exactly ASL (American Sign Language) but they are adapted from ASL.

Content
2125 images
Image size is 300x300
The first letter indicates the label (e.g. A-1554389290.JPG)
"J" is missing

Team Members

Implemented with great efforts of

• Fuad Aghazada
• K. Ayberk Tecimer
• Emre Sülün
• Adahan Yalçınkaya
• Eray Şahin

Demo link