<![CDATA[In this study, we used a commonly used object detection algorithm to classify sign language gestures, namely BISINDO or Indonesian Sign Language. The process of learning sign language is still limited, especially with the use of traditional methods such as direct conversation or using a dictionary. However, there are still obstacles with this approach, for example, some students have difficulty interpreting what they see in the dictionary. Therefore, this study aims to overcome this problem by using a real-time image classification model. The dataset used in this study was collected by the researchers themselves, with a total of 520 images consisting of 26 classes of BISINDO alphabet gestures. We also used transfer learning in this study to utilize the pre-trained SSDMobileNet V2 architecture. Using the COCO evaluation metric, the results show that this model achieved 94% mean average precision, 91% average precision, and 85% recall. This model can also classify sign language gestures in real-time.]]>
Copyrights © 2025
