L. Abou-Abbas,
C. Tadj, H. F. Alaie
Detection
of cry sounds is generally an important
pre-processing step for various applications involving cry analysis
such as diagnostic systems electronic monitoring systems,emotion
detection and robotics for baby caregivers.Given its complexity, an
automatic cry segmentation system is a rather challenging topic.A new
framework for automatic cry sound segmentation for application in a
cry-based diagnostic system has been proposed.We studied the
contribution of various additional time-frequency domain features to
increase the robustness of a GMM/HMM-based cry
segmentation system in noisy environments.We introduced a fully
automated segmentation algorithm to extract cry sound
components,audible expiration&inspiration,based on two
approaches:statistical analysis based on GMM/HMM classifiers and a
post-processing method based on intensity,zero crossing rate,and
fundamental frequency feature extraction.The main focus of this paper
is to extend the systems developed in our previous works to include a
post-processing stage with a set of corrective and enhancing tools to
improve the classification performance.This full approach allows us to
precisely determine the start&end points of the expiratory and
inspiratory components of a cry signal, EXP and INSV in any given sound
signal.Experimental results have indicated the effectiveness of the
proposed solution. EXP & INSV detection rates of approximately 94.29% and 92.16% respectively were achieved by
applying a 10-fold cross-validation technique to avoid over-fitting.