Lim, Boon Cheong (2019) Hand Gestures Recognition Using Deep Neural Network. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)
![]() |
Text (24)
Hand Gestures Recognition Using Deep Neural Network.pdf - Submitted Version Download (936kB) |
Abstract
Gestures are another type of communication tool which can be used to express idea, thought and feeling. It is usually used by those who have hearing and speech disability as mother language. This project was about implementation of a deep learning-based recognitions system for 30 MSL static hand gestures and evaluation of the performance of the system designed in term of recognition accuracy. This project was designed for those users who using the recognition system in front of an USB camera and the system designed was capable to recognize 30 MSL hand gesture sonly. The recognition system designed was made up by two hierarchical CNN architectures, in which YOLOV2 model was used for hand detection while MobileNet pretrained model was used for gestures classification ofthis project. Throughout this project, the hand gestures recognition system designed achieved 98 % average testing accuracy on self-generated testing dataset. Given the limitations of the datasets and the encouraging results achieved, a fully generalizable translator for all 30 MSL static hand gestures can be produced with further research and inclusion of more dataset for MobileNet classification training.
Item Type: | Final Year Project (Project Report) |
---|---|
Uncontrolled Keywords: | Gesture, Human-computer interaction |
Subjects: | T Technology > T Technology (General) T Technology > TA Engineering (General). Civil engineering (General) |
Divisions: | Library > Final Year Project > FKEKK |
Depositing User: | Norfaradilla Idayu Ab. Ghafar |
Date Deposited: | 13 Jul 2020 02:50 |
Last Modified: | 28 Jul 2020 07:23 |
URI: | http://digitalcollection.utem.edu.my/id/eprint/24457 |
Actions (login required)
![]() |
View Item |