Classifying Hand Gestures for People with Disabilities Utilizing the MobileNetV2 Model

Approximately 466 million people worldwide experience deafness or hearing impairment. These individuals face the challenge of connecting with others., acquiring new knowledge., and participating in various activities. gesture language acts as a crucial bridge., facilitating communication between them and the broader world. However., automating the recognition of sign language poses numerous challenges. To address this., we utilizing the MobileNetV2 Model. Our approach involved utilizing MobileNetV2 Model with dataset that represents the ASL Alphabet., encompassing 24 classes corresponding to English gesture language letters., (excluding the letters J and Z., which necessitate movement). The training dataset comprises 27.,455 entries. cases., while the test dataset comprises 7.,172 cases. In our training process., 80% of the dataset was designated for training., with the remaining 20% utilized for testing. The results of our proposed model were highly promising., achieving a remarkable accuracy rate of 99.9% and 100% validation accuracy. This success indicates the potential effectiveness of our model in automatically recognizing American Sign Language gestures., thereby contributing to improved communication and accessibility for individuals with hearing challenges.