Breaking the silence: empowering the mute-deaf community through automatic sign language decoding

Area of Science:

• Computer Vision
• Artificial Intelligence
• Human-Computer Interaction

Background:

• Communication barriers exist for the deaf-mute community in Egypt.
• Existing solutions for sign language recognition are limited.
• Need for accessible technology to bridge communication gaps.

Purpose of the Study:

• To develop an automatic sign language recognition system.
• To improve the quality of life for the deaf-mute community in Egypt.
• To convert right-hand gestures into audible sounds or displayed text.

Main Methods:

• Utilized a convolutional neural network (CNN) model for gesture recognition.
• Trained the CNN on right-hand gestures captured via a web camera.
• Created a custom dataset with volunteer participation for model training and validation.

Main Results:

• Achieved an average accuracy of 99.65% in recognizing right-hand gestures.
• Attained a high precision value of 95.11%.
• Successfully distinguished between similar gestures of different alphabets.

Conclusions:

• The proposed system offers a viable solution for automatic sign language recognition in Egypt.
• Accurate gesture identification enhances communication for the deaf-mute community.
• The technology promotes inclusivity and accessibility, improving quality of life.