Journal Article

·2018

The Effect of Convolutional Neural Network Parameters on Sign Language Recognition

Osman Oğuzhan Savaş YTU , Tülay Yıldırım YTU

Abstract

The sign language has been used for many years by the hearing and speech impaired in their communication; It is a visual language consisting of hand, face and body movements. Sign language studies have been started to provide communication between disabled citizens and other persons in this way. These studies, which started with sensor gloves, continued with computerized vision methods after 90s due to various advantages. In this study, the sign language movements were determined with the model of Convolutional Neural Network (ESA) which is one of the deep learning models and the effect of ESA parameters on the system performance ratio was examined. For this purpose, data set containing letters and numbers prepared for American Sign Language (ASL) was used.

Keywords

Sign language Convolutional neural network Computer science Sign (mathematics) American Sign Language Speech recognition Set (abstract data type) Artificial intelligence Face (sociological concept) Natural language processing Linguistics Mathematics

Subject Areas

Hand Gesture Recognition Systems ·Human-Computer Interaction ·Physical Sciences
Gait Recognition and Analysis ·Biomedical Engineering ·Physical Sciences
Human Pose and Action Recognition ·Computer Vision and Pattern Recognition ·Physical Sciences

Citations by Year