Hierarchical voting classification scheme for improving visual sign language recognition

Authors:
Liang-Guo Zhang;Xilin Chen;Chunli Wang;Wen Gao
Affiliations:
Chinese Academy of Sciences, Beijing, China and Harbin Institute of Technology, Harbin, China;Chinese Academy of Sciences, Beijing, China and Harbin Institute of Technology, Harbin, China;Chinese Academy of Sciences, Beijing, China;Chinese Academy of Sciences, Beijing, China and Harbin Institute of Technology, Harbin, China
Venue:
Proceedings of the 13th annual ACM international conference on Multimedia
Year:
2005

Citing 5
Cited 0

Real-Time American Sign Language Recognition Using Desk and Wearable Computer Based Video

IEEE Transactions on Pattern Analysis and Machine Intelligence
A framework for recognizing the simultaneous aspects of American sign language

Computer Vision and Image Understanding - Modeling people toward vision-based underatanding of a person's shape, appearance, and movement
An Empirical Comparison of Voting Classification Algorithms: Bagging, Boosting, and Variants

Machine Learning
A Real-Time Continuous Gesture Recognition System for Sign Language

FG '98 Proceedings of the 3rd. International Conference on Face & Gesture Recognition
A vision-based sign language recognition system using tied-mixture density HMM

Proceedings of the 6th international conference on Multimodal interfaces

Quantified Score

Hi-index	0.00

Visualization

Abstract

As one of the important research areas of multimodal interaction,sign language recognition (SLR) has attracted increasing interest.In SLR, especially on medium or large vocabulary, it is usuallydifficult or impractical to collect enough training data. Thus, howto improve the recognition on the limited training samples is asignificant issue. In this paper, a simple but effectivehierarchical voting classification (HVC) scheme for improvingvisual SLR, which makes efficient use of limited training data, isproposed. The key idea of HVC scheme is similar to but not the sameas Bagging technique. Firstly, it constructs several training setsfrom the original training set in a combinatorial fashion togenerate the corresponding continuous hidden Markov models (CHMM)ensemble. Then, it determines the ensemble output by appropriatelocal voting strategy. Finally, it obtains the final recognitionresult by the global voting. Experimental results show that the HVCscheme outperforms the conventional single CHMM approach in termsof recognition accuracy on the limited training data.