An HMM-based threshold model approach for gesture recognition

• Published in: IEEE transactions on pattern analysis and machine intelligence Vol. 21; no. 10; pp. 961 - 973
• Main Authors: Hyeon-Kyu Lee, Kim, J.H.
• Format: Journal Article
• Language: English
• Published: IEEE 01.10.1999
• Subjects: Construction; Entropy; Handles; Hidden Markov models; Impedance matching; Intelligence; Mathematical models; Pattern analysis; Pattern matching; Pattern recognition; Probability distribution; Recognition; Shape; Spatiotemporal phenomena; Speech recognition; Switches; Thresholds

A new method is developed using the hidden Markov model (HMM) based technique. To handle nongesture patterns, we introduce the concept of a threshold model that calculates the likelihood threshold of an input pattern and provides a confirmation mechanism for the provisionally matched gesture pattern...