Deteksi Objek Bahasa Isyarat Huruf Bisindo Menggunakan SSD-Mobilenet

Gunawan Abdillah (Universitas Jenderal Achmad Yani, Indonesia)
Ridwan Ilyas (Universitas Jenderal Achmad Yani, Indonesia)

Kesatria : Jurnal Penerapan Sistem Informasi (Komputer dan Manajemen)
Vol 5, No 1 (2024): Edisi Januari

Automatic sign language recognition systems help the hearing-impaired community communicate better. The pure sign language system developed by deaf Indonesians is called BISINDO. BISINDO is used by deaf friends based on their knowledge of their environment. SSD is an abbreviation for Single Shot MultiBox Detector, a method of detecting objects in images using a neural network in one stage. With SSD, objects of any size and shape can be identified easily and accurately without the need for object suggestions or complex resampling steps. This research uses SSD Mobile Net to identify bisindo sign language for the Letter category. The evaluation results show that the best model is SSD Mobilenet V1 FPN 640 x 640.