Anthropomorphic robotic hand for basic Malaysian sign language /
The community of speech/hearing impaired individuals makes up a noticeable segment of the society that drew the researchers' attention in the last several decades, especially with the development in the field of social robotics, and the sign language robots hold promise for this issue. The tran...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
Kuala Lumpur :
Kulliyyah of Engineering, International Islamic University Malaysia,
2018
|
| Subjects: | |
| Online Access: | http://studentrepo.iium.edu.my/handle/123456789/4341 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The community of speech/hearing impaired individuals makes up a noticeable segment of the society that drew the researchers' attention in the last several decades, especially with the development in the field of social robotics, and the sign language robots hold promise for this issue. The translation from spoken language to sign language consists of two main parts which are the speech recognition system and the performance of the anthropomorphic robotic hand. Although the integration of these two aspects is the focal point of performing a more realistic sign language, the previous works have overlooked the integration between both sides. Therefore, the existing systems are mostly redundant and less flexible. Accordingly, this study aims at developing an integrated an anthropomorphic hand for basic Malaysian Sign Language (MSL) that interprets the uttered speech to the signed motion. The study proposes a speech to sign language translation system that recognizes the human natural speech and converts it into signed robotic hand motions that can be understood by hearing-impaired individuals. In this study, the speech signals have been analysed by means of speech recognition system using Microsoft Visual Studio platform, C# programming language. An anthropomorphic robotic hand has been designed. The result shows that the accuracy of the speech recognition system obtained by Confusion Matrix, is 92.69%, which is sufficient for integrating the speech system with the robot motion. The proposed anthropomorphic robotic hand for Malaysian Sign Language has been tested with the Malaysian Federation of Deaf, and the results have proven the capability of the robot to perform basic signs selected from Malaysian Sign Language (MSL) with an accuracy of 91.7%. This work, to our knowledge, is the only work that considers all aspects of the sign language robotic translation from integrating the right speech recognition system to the mechanical performance by the robot. It is the only work in performing the MSL through an anthropomorphic robotic hand. It is believed that this study will bridge the gap of communication between normal and hearing-impaired individuals in Malaysia, at first point, and throughout the world. |
|---|---|
| Physical Description: | xvi, 102 leaves : colour illustrations ; 30cm. |
| Bibliography: | Includes bibliographical references (leaves 85-94). |