Dynamic Hand Gesture Recognition Based on Randomized Self-Organizing Map Algorithm

Authors:
Tarek El. Tobely;Yuichiro Yoshiki;Ryuichi Tsuda;Naoyuki Tsuruta;Makoto Amamiya
Affiliations:
-;-;-;-;-
Venue:
ALT '00 Proceedings of the 11th International Conference on Algorithmic Learning Theory
Year:
2000

Citing 4
Cited 1

Digital neural networks

Digital neural networks
Recognizing hand gestures

ECCV '94 Proceedings of the third European conference on Computer vision (vol. 1)
Artificial intelligence: a modern approach

Artificial intelligence: a modern approach
Self-organizing maps

Self-organizing maps

A comparison of 3D hand gesture recognition using dynamic time warping

Proceedings of the 27th Conference on Image and Vision Computing New Zealand

Quantified Score

Hi-index	0.00

Visualization

Abstract

Gesture recognition is an appealing tool for natural interface with computers especially for physically impaired persons. In this paper, it is proposed to use Self-Organized Map (SOM) to recognize the posture images of hand gestures. Since the competition algorithm of SOM allows alleviating many difficulties associated with gesture recognition. However, it is required to reduce the recognition time of one image in SOM network to the range of normal video camera rates, this permits the network to accept dynamic input images and to perform on-line recognition for hand gestures. To achieve this, the Randomized Self-Organizing Map algorithm (RSOM) is proposed as a new recognition algorithm for SOM. With RSOM algorithm, the recognition time of one image reduced to 12.4 % of the normal SOM competition algorithm with 100 % accuracy and allowed the network to recognize images within the range of normal video rates. The experimental results to recognize six dynamic hand gestures using RSOM algorithm is presented.