Sign Language Detector using Python

This is a Sign Language Detector using Python. The sign language used is not the usual ASL (American Sign Language), but SIBI (Sistem Bahasa Isyarat Indonesia), which is the official sign language in Indonesia.

Technicals

• The dataset was created manually from scratch with data augmentation. There are a total of 200 data for each letter.
• The model used was a traditional Machine Learning model, which is Random Forest.

Requirements

• opencv-python==4.7.0.68
• mediapipe==0.10.21
• scikit-learn==1.3.1

How To Run the System

1. Clone the directory
   In the terminal of your directory:

   git clone https://github.com/Hansen-Susilo/Sign-Language-Detection.git
   

2. Install the requirements
   

   pip install -r requirements.txt
   

   Note: Python version must be 3.12 or lower. This code will not work in 3.13 due to the mediapipe dependency not supported
3. Open the file "inference_classifier.py"
4. Run the file

   python inference_classifier.py
   

References

• https://www.youtube.com/watch?v=MJCSjXepaAM
• https://lbifib.ui.ac.id/index.php/id/blog/artikel/perbedaan-bahasa-isyarat-bisindo-vs-sibi