An HMM-based threshold model approach for gesture recognition

TLDR

The threshold model is a weak model for all trained gestures, having a lower likelihood than the dedicated gesture model for each gesture. The study develops an HMM‑based threshold model to handle nongesture patterns and confirm provisional gesture matches. The method uses the likelihood from the threshold model as an adaptive threshold, reduces its large state space by merging states with similar probability distributions using relative entropy, and selects the appropriate gesture model. Experimental results demonstrate 93.14 % reliability in extracting trained gestures from continuous hand motion.

Abstract

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 patterns. The threshold model is a weak model for all trained gestures in the sense that its likelihood is smaller than that of the dedicated gesture model for a given gesture. Consequently, the likelihood can be used as an adaptive threshold for selecting proper gesture model. It has, however, a large number of states and needs to be reduced because the threshold model is constructed by collecting the states of all gesture models in the system. To overcome this problem, the states with similar probability distributions are merged, utilizing the relative entropy measure. Experimental results show that the proposed method can successfully extract trained gestures from continuous hand motion with 93.14% reliability.