We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Intelligent Gesture Recognition System for Deaf People by using CNN and IoT.
- Authors
Abualkishik, Abdallah; Alzyadat, Wael; Al Share, Marwan; Al-Khaifi, Sara; Nazari, Mojtaba
- Abstract
Communication with the hearing impaired is a great challenge in the society today. Sign language is a significant communication method between deaf persons and their societies. However deaf and dumb people have an issue in this field as sign language may cause a lot of misunderstanding. Today, many new technologies are becoming more applicable and cheaper. The intelligent algorithms and techniques are improved and becomes more accurate. Gesture recognition is one of the key technologies that facilitate many people life. Translating signs language into text and speech is an attractive field for researchers from past few years. However, it did not receive adequate interest in Sultanate of Oman. The proposed recognition system utilizes the Convolution neural network (CNN) method. It aims to convert the dynamic deaf signs from live video to text and speech using raspberry pi device and normal camera. The dataset for this project was created by the researchers. It contains 62000 (64x64 pixel) images of the 30 letter of Alphabets and 1 Word. Each pattern has 2000 images that are divided into 1750 images for training and 250 images for testing. The proposed system achieved 99.8 % accuracy.
- Subjects
OMAN; DEAF people; CONVOLUTIONAL neural networks; GESTURE; SIGN language; HEARING impaired; HIDDEN Markov models; COCHLEAR implants
- Publication
International Journal of Advances in Soft Computing & Its Applications, 2023, Vol 15, Issue 1, p144
- ISSN
2710-1274
- Publication type
Article
- DOI
10.15849/IJASCA.230320.10