206
Kayumov O.A., Sayfullayeva I.A., Rustamov I.R.
UZSL creating an interactive intellectual
e-learning “uzsl online learning platform” for teaching uzbek sign language
UZSL CREATING AN INTERACTIVE INTELLECTUAL E-LEARNING “UZSL ONLINE
LEARNING PLATFORM” FOR TEACHING UZBEK SIGN LANGUAGE
Kayumov Oybek Achilovich
Head of the Computer Science and Programming Department, Jizzakh branch of the
National University of Uzbekistan named after Mirzo Ulugbek, Jizzakh, Uzbekistan.
E-mail: oybekuzonlined3@gmail.com
Tel: 94 573-31-83
Sayfullayeva Iroda Anvar qizi
Student of computer science and programming technologies, Jizzakh branch of the
National University of Uzbekistan named after Mirzo Ulugbek, Jizzakh, Uzbekistan.
Rustamov Ibrohim Raxmat oʻgʻli
Student of computer science and programming technologies, Jizzakh branch of the
National University of Uzbekistan named after Mirzo Ulugbek, Jizzakh, Uzbekistan.
Abstract:
The aim of this study is to design a platform that can recognize the dactyl
alphabet, a form of sign language based on Latin script and hand movements, and convert it
into text. Additionally, the platform seeks to promote sign language education among all
members of society. This technology can help reduce the intensive speech development
period for deaf and hard-of-hearing children, promote sign language use in families and
public services, and serve as a basic didactic resource for teachers and parents. In
educational settings, the platform can aid deaf and hard-of-hearing students in developing
a thematic vocabulary to enhance their speech reserve. The purpose of sign language
technologies is to bridge the communication gap between the deaf and hard of hearing
community and the rest of society. Real-time sign language recognition (UzSLP) is a
cutting-edge platform that facilitates communication between individuals who are deaf or
hard of hearing and others. Our research employs transfer learning to enable hand-
tracking-based sign language recognition using a Convolutional Neural Network (CNN)
model for image classification. This approach allows for fast computation and can be
perform in real-time on the platform.
