The impact of geometric and motion features on sign language translators
Malaysian Sign Language (MSL) recognition system is a choice of augmenting communication between the hearing-impaired and hearing communities in Malaysia. Automatic translators can play an important role as alternative communication method for the hearing people to understand the hearing impaired...
| Main Authors: | , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English |
| Published: |
WSEAS E-Library
2015
|
| Subjects: | |
| Online Access: | http://irep.iium.edu.my/46566/ http://irep.iium.edu.my/46566/ http://irep.iium.edu.my/46566/1/Impact_of_geometric_of_motion_features.pdf |
| Summary: | Malaysian Sign Language (MSL) recognition system is a choice of augmenting communication
between the hearing-impaired and hearing communities in Malaysia. Automatic translators can play an
important role as alternative communication method for the hearing people to understand the hearing impaired
ones. Automatic Translation using bare hands with natural gesture signing is a challenge in the field of machine
learning. Researchers have used electronic and coloured gloves to solve mainly three issues during the preprocessing
steps before the singings’ recognition stage. First issue is to differentiate the two hands from other
objects. This is referred to as hand detection. The second issue is to describe the detected hand and its motion
trajectory in very descriptive details which is referred to as feature extraction stage. The third issue is to find the
starting and ending duration of the sign (transitions between signs). This paper focuses on the second issue, the
feature extraction by studying the impact of the vector dimensions of the features. At the same time, signs with
similar attributes have been chosen to highlight the importance of features’ extraction stage. The study also
includes Hidden Markov Model (HMM) capability to differentiate between signs which have similar attributes. |
|---|