KLASIFIKASI HURUF DAN ANGKA PADA BAHASA ISYARAT MENGGUNAKAN METODE CONVOLUTIONAL NEURAL NETWORK (CNN)
Total View This Week0

Institusion

Universitas Sriwijaya

Author

ARGABZI, MUHAMMAD
Samsuryadi, Samsuryadi
Rachmatullah, Muhammad Naufal

Subject

Q334-342 Computer science. Artificial intelligence. Algorithms. Robotics. Automation. 

Datestamp

2023-05-22 04:11:27 

Abstract :
Sign language is one of the communication media that can be used by deaf people, however the use of sign language is not only be utilized by the one who disabled on it, but also can be learned and used by normal people. Classification of signals uses a convolutional neural network (CNN) algorithm which is capable of obtaining important features from each image without human assistance. In addition, the CNN algorithm is more efficient when compared to other neural network algorithms, especially for memory and complexity. AlexNet is an appropriate architecture to be applied in this research. This classification uses 34 classes, providing 8 test scenarios. The highest classification result of the 8 scenarios is 98%. The CNN algorithm can perform sign language classification with high accuracy. 

File :
RAMA_55201_09021381823143.pdf
RAMA_55201_09021381823143_TURNITIN.pdf
RAMA_55201_09021381823143_0004027101_0001129204_01_front_ref.pdf
RAMA_55201_09021381823143_0004027101_0001129204_02.pdf
RAMA_55201_09021381823143_0004027101_0001129204_03.pdf
RAMA_55201_09021381823143_0004027101_0001129204_04.pdf
RAMA_55201_09021381823143_0004027101_0001129204_05.pdf
RAMA_55201_09021381823143_0004027101_0001129204_06.pdf
RAMA_55201_09021381823143_0004027101_0001129204_07_ref.pdf
RAMA_55201_09021381823143_0004027101_0001129204_08_lamp.pdf