IMPLEMENTASI DEEP LEARNING UNTUK MENGKLASIFIKASI ABJAD BAHASA ISYARAT MENGGUNAKAN ALGORITMA CONVOLUTION NEURAL NETWORK
Total View This Week2
Total View This Week2
Institusion
STMIK Bumi Gora
Author
Karnaen, L Yuda Rahmani
Subject
T Technology (General)
Datestamp
2021-09-04 02:02:24
Abstract :
Sign Language is the language used by people with special needs to communicate manually, body language, and lip movements instead of using sounds and voices to communicate. The use of sign language in communicating between deaf and speech impaired people with normal people is a social problem that often occurs. SIBI or Indonesian Sign System is a language issued by the Ministry of Education and Culture and is applied in special school lessons (SLB). Convolutional Neural Network (CNN) is a deep learning (DL) method that can be used to detect and recognize an object in a digital image. CNN's ability is claimed as the best model to solve object detection and object recognition problems. The training dataset uses CNN architecture with Adam as the optimizer. The training was carried out twice, namely by using 10 iterations and 20 iterations. The accuracy of the model that has been made reaches 99%, while the accuracy of identification based on the webcam reaches 79.31% in bright areas while in dim areas it reaches 65.51%.
Sign Language is the language used by people with special needs to communicate manually, body language, and lip movements instead of using sounds and voices to communicate. The use of sign language in communicating between deaf and speech impaired people with normal people is a social problem that often occurs. SIBI or Indonesian Sign System is a language issued by the Ministry of Education and Culture and is applied in special school lessons (SLB). Convolutional Neural Network (CNN) is a deep learning (DL) method that can be used to detect and recognize an object in a digital image. CNN's ability is claimed as the best model to solve object detection and object recognition problems. The training dataset uses CNN architecture with Adam as the optimizer. The training was carried out twice, namely by using 10 iterations and 20 iterations. The accuracy of the model that has been made reaches 99%, while the accuracy of identification based on the webcam reaches 79.31% in bright areas while in dim areas it reaches 65.51%.
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
