International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 03 | Mar 2024
p-ISSN: 2395-0072
www.irjet.net
A Survey on Sign Language Recognition Ananya Jain Chattahoochee High School, 5230 Taylor Rd, Johns Creek, GA 30022, United States. Email ID: ananyajain0719@gmail.com
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract – Sign language plays a crucial role in facilitating communication for the deaf and hearing-impaired people in society. According to the World Health Organization, around 466 million people rely on sign language for approximately 300 distinct languages, while 56% of sign language remains dominated by English. Thus, developing accurate sign language recognition (SLR) models remains pivotal for a large section of society. In this work, we present a comprehensive overview of the various algorithms developed for sign language recognition. We broadly classify the different types of SLR based on the number of hands employed for hand gestures, vaying types of inputs for SLR models, etc. Further, we present a detailed examination of sensor-based SLR models that employ external objects like gloves, and vision-based SLR models that focus only on hand gestures images and videos. Recently, artificial intelligence has appeared as a promising approach for dramatically improving the accuracy of SLR models. Thus, we present a critical analysis of machine learning models employed for SLR, such as k-nearest neighbor, support vector machines, etc. Moreover, deep learning has enabled the prediction of sign languages with very high accuracy surpassing all the state-of-the-art methods. Thus, we present a detailed analysis of the state-of-the-art deep learning models employed for SLR, such as convolutional neural network and long short-term memory models.
Fig-1: The global distribution of various sign languages. into computer commands or just plain text that other people can comprehend, modern devices are either cameracompatible or include built-in cameras. However, since speech recognition systems lag sign language recognition (SLR) systems by about 30 years, SLR system development is essential to ensure that the public welfare benefits from the technological advancements even for those in the hearingimpaired community.
Key Words: Sign language recognition, Gestures, Machine learning, Deep learning, and Artificial intelligence.
In recent years, one of the most promising approaches has emerged: artificial intelligence (AI). Due to significant technological advancements, artificial intelligence (AI) has drawn countless scientific and practical applications that are transforming every sector. Fig. 2 enumerates a few domains where AI is operating. One such application that has attracted a lot of interest in the last 10 years is computer vision (CV). With the use of machine learning (ML) and deep learning (DL) models, CV processes digital photos and videos as input and produces meaningful conclusions. Throughout the world, computer vision is quickly changing a wide range of sectors, including robotics, finance, agriculture, healthcare, auto-mobiles (self-driving autonomous automobiles), security systems, and many more [2]. Realtime, high-speed computing resources with minimal power consumption are necessary for the methods used to tackle CV challenges.
1.INTRODUCTION The World Health Organisation estimates that 466 million people worldwide are deaf, including 432 million adults and 34 million children. Sign Language enables the deaf community to communicate with one another and with the outside world. Sign language assists in breaking down barriers between the deaf population and the rest of the world [1]. There are around 300 distinct sign languages in use across the world, which vary by country. In Fig. 1, we depict the global popularity of several sign languages (in percentages). Recently, the advent of voice-activated personal digital assistants (PDAs) such as Apple's Siri and OK Google has revolutionized the way people live. However, not many changes have been made to support individuals with hearing impairments, so these technological advances are not readily available to them. In order to help translate hand gestures
© 2024, IRJET
|
Impact Factor value: 8.226
In this paper, we address the crucial computer vision problem of sign language recognition (SLR), which affects a significant portion of the general population. We perform a thorough literature assessment of the most recent state-of-
|
ISO 9001:2008 Certified Journal
|
Page 909