Hand sign recognition from intensity image sequences with complex backgrounds
FG '96 Proceedings of the 2nd International Conference on Automatic Face and Gesture Recognition (FG '96)
A Real-Time Continuous Gesture Recognition System for Sign Language
FG '98 Proceedings of the 3rd. International Conference on Face & Gesture Recognition
KernTune: self-tuning Linux kernel performance using support vector machines
Proceedings of the 2007 annual research conference of the South African institute of computer scientists and information technologists on IT research in developing countries
Hi-index | 0.00 |
The South African Sign Language research group at the University of the Western Cape has created several systems to recognize Sign Language gestures using single parameters. Research has shown that five parameters are required to recognize any sign language gesture: hand shape, location, orientation and motion, as well as facial expressions. Using a single parameter can cause conflicts in recognition of signs that are similarly signed. This paper pioneers research at the group towards combining multiple parameters to better distinguish between similar signs. This eventually aims to enable the recognition of a large SASL vocabulary. The proposed methodology combines hand location and hand shape recognition into one combined recognition system. The recognition approach is applied to 12 SASL signs that consist of six pairs of signs with the same hand shape performed at two different locations. It is shown that the approach is able to achieve a high average recognition accuracy of 79% across all signs and distinguish between the signs effectively. It is also shown to be robust to variations in test subjects.