Sindhi Alphabetical Sign Language Recognition Using Machine Learning

Abstract

Sign language is an essential communication medium for deaf and hard-of-hearing individuals. While significant progress has been made in recognizing widely used sign languages, regional languages like Sindhi Sign Language (SSL) remain underexplored. This research proposes a machine learning-based system for recognizing Sindhi alphabetical signs from hand gesture images.A dataset consisting of 2,080 images representing 52 Sindhi alphabets was created using data collected from over 40 participants. Pre-processing techniques such as grayscale conversion, resizing, normalization, and data augmentation were applied to improve model performance. Feature extraction was performed using the Histogram of Oriented Gradients (HOG) method, followed by classification using Support Vector Machine (SVM). The model was evaluated using performance metrics such as accuracy, precision, recall, and F1-score. Results show that the system achieves reliable performance in recognizing Sindhi sign gestures, demonstrating the potential of machine learning in assistive communication technologies.