<![CDATA[Real-time Indonesian sign language recognition faces several challenges, the most prominent of which is the diversity of hand gestures and different expressions. Often, automated systems face difficulties in interpreting these highly variable gestures, which in turn results in decreased recognition accuracy and efficiency. To improve the performance of sign language recognition in this study, the SP-Tree method is proposed, which utilizes a spatial tree structure to group hand gesture data based on spatial and temporal features. This allows for a faster and more accurate sign language recognition process. It is expected that this technique can accelerate sign language recognition with a high level of accuracy and real-time response, which is very important for everyday applications. We used a public dataset covering various hand gestures in Indonesian sign language to test this technique. The results showed that the SP-Tree method had an accuracy of 92 percent, an F1 score of 0.90, and a feature loss of 0.08. Compared with existing conventional sign language recognition techniques, these figures show significant improvements. The results indicate that the SP-Tree method is an effective way to identify Indonesian signs in real-time. This method has the advantage of being able to interpret and group hand gestures more precisely and efficiently, improving the interaction between the user and the system. We hope that this research will help develop assistive technology for people with hearing disabilities. In the future, it will also provide opportunities to use this technique in other sign languages.]]>
Copyrights © 2025
