Cry Segments Research Articles

Convolutional Neural Networks for Audio-Based Continuous Infant Cry Monitoring at Home

Cry is an important signal in early infancy for parents to understand needs of their baby and thereby to provide timely parenting/soothing or to be reassured. Thanks to the recent advancement of signal processing, deep learning, and internet-of-things technologies, smart baby monitors with a microphone and/or a video camera have attracted a lot of attention to be used in a baby room to assist parental activities. In this paper, we propose a two-step approach to detect infant cries automatically with continuous audio signals. We first identify and remove the segments without clear sounds (background noise) using a volume-based thresholding algorithm, followed by convolutional neural network (CNN) models to further detect infant cries. The CNN operates on the log linear-scale filterbank energies of audio signals to extract features for cry detection. In this study, a large set of audio data (151.8 hours) collected from five infants in home settings were included. Our proposed approach achieved a mean accuracy of 98.6% in identifying background noise (with only 2 out of 3209 cry segments missed) and a mean accuracy of 92.2% in detecting cries from other non-background sounds.

Automatic infant cry analysis for the identification of qualitative features to help opportune diagnosis

In the infant cry analysis, the identification of qualitative features is of great importance, because this provides relevant information to differentiate between normal and pathological cries, which makes important their identification. Qualitative infant cry analysis has been done until now by medical personal through visual inspection of spectrograms and by the auditory study of the cry recordings. In this way, the success of the process depends on the subjective perception of the inspector besides being a very slow task. The information extracted from the perceptive observation of the crying waves recordings is then used as a help to emit diagnosis. With the idea of helping to make the whole process easier and faster we are developing a method to automatically identify, measure and highlight selected qualitative features in infant cry recordings. The processing of this identifier starts with the automatic discovery of infant cry units, which is performed by the use of a threshold applied to the energy of the signal along with another threshold applied to eliminate inspiratory cry segments, when not needed. From all the detected cry units, the process automatically identifies melodic shape, shifts, glides and noise concentration. In this work, we present, besides a quick review of related works, and a description of the perceptive analysis to help diagnosis, the process implementation, some experiments as well as the experimental results obtained.

Acoustic characteristics of crying in infantile laryngomalacia

The purpose of this study was to add to the extant data base of acoustic cry studies by profiling the condition of laryngomalacia. We hypothesized that the acoustic characteristics of crying produced by an infant with laryngomalacia would differ compared to previously reported cry data for normal infants. An entire episode of crying was audio recorded and acoustically analyzed for the occurrence of expiratory and inspiratory cry segments, as well as the long-time average spectral (LTAS) characteristics. Results obtained for the infant were found to be considerably different from what has been previously reported for normal infants. The overall duration of the infant's crying episode was longer, with proportionately fewer expiratory phonations and more inspiratory phonations compared to normal infants. The LTAS results were reflective of aperiodic components in the glottal source spectrum. Collectively, the infant's unusual crying aspects were not limited solely to those acoustic features resulting from a prolapse of supraglottic soft tissue, and therefore provide new insight into the vocal fold vibratory behavior characterizing infantile laryngomalacia.

Acoustic Characteristics of Naturally Occurring Cries of Infants with "Colic"

Although infantile colic has long been defined by a perceived excessive amount of crying, acoustic attributes of the cry sound may also contribute to perceptions that this early social behavior is excessive or problematic. From an original sample of 76 infants (38 infants referred to physicians for problematic crying, or "colic," and 38 pair-matched comparison infants), 48 infants who produced naturally occurring cry bouts both before and after an evening feeding were studied: 11 infants with Wessel's colic, 15 infants with non-Wessel's colic, and 22 comparison infants. Standard and vociferous cry segments were selected from up to 2 min of tape-recorded crying for spectrum analysis. Vociferous cry segments had a longer duration, a higher fundamental frequency, and a greater percentage of dysphonation than did standard segments. No differences between infant groups were found in cries before feeding. After feeding, infants who were problematic criers, independent of Wessel's criteria, showed a greater percentage of dysphonation in the vociferous cry segment than did comparison infants. This finding resulted from a decrease in dysphonation in the cries of comparison infants after feeding and an increase in those of infants with non-Wessel's colic. The dominant frequency also increased after feeding in the vociferous cries of infants with Wessel's colic, resulting in these infants having higher-pitched cries after feeding than infants in the other 2 groups. Results indicate that infants who are perceived to have problematic crying have objectively different acoustic features in their cry sounds that are particularly aversive, and that complaints about excessive crying cannot be accounted for simply on the basis of reporting bias in overly concerned or emotionally labile parents.

The long-time average spectra of newborn infant crying

The majority of research pertaining to acoustic analysis of infant crying is based on the collection of a limited number of cry segments. Far less research is based on examination of whole crying episodes. Long-time average spectrum (LTAS) analysis offers a quantitative approach to inferring the vocal source as derived from whole crying episodes. The purpose of this study was to quantitatively examine a variety of LTAS characteristics (e.g., energy concentrations, spectral tilt) in the crying episodes of ten normal full-term neonates and ten at-risk preterm neonates. Results of this analysis indicated that some but not all of the LTAS features were discriminating of normal versus at-risk infants. The discussion focuses on the application of cry analysis in the evaluation of laryngeal behavior and its potential usefulness in determining an infant’s health status.

Application of an acoustic cry template to evaluate at-risk newborns: preliminary findings.

The durational characteristics of crying episodes produced among two groups of at-risk newborn infants are reported. The groups included preterm infants with diagnosed respiratory problems and preterm infants with diagnosed neurological problems. Each infant's complete crying episode was audio-recorded and acoustically analyzed for overall occurrence of expiratory cry segments, inspiratory cry segments, and nonphonatory 'pause' segments. Remarkable similarity was found across risk groups with regard to the distributional occurrence of the cry features. Results obtained for the at-risk infants were highly similar to those observed for normal full-term newborns. These preliminary findings are taken to suggest that temporal features of newborn crying activity reflect a basic organizational response of the nervous system which is common to both normal and at-risk infants.

An acoustic template of newborn infant crying.

The durational characteristics of 10 normal infant's complete crying episodes are reported. The cries of each child were audiorecorded at 2 days postpartum and acoustically analyzed for overall occurrence of expiratory cry segments, inspiratory cry segments, and nonphonatory 'pause' segments. In general, remarkable similarity was found across infants with regard to the distributional occurrence of the cry features. Discussion focuses on the organization of newborn infant crying behavior within the milieu of a crying episode, as well as the diagnostic implications of these findings.

4977812 Accompaniment line principal tone determination system

In the infant cry analysis, the identification of qualitative features is of great importance, because this provides relevant information to differentiate between normal and pathological cries, which makes important their identification. Qualitative infant cry analysis has been done until now by medical personal through visual inspection of spectrograms and by the auditory study of the cry recordings. In this way, the success of the process depends on the subjective perception of the inspector besides being a very slow task. The information extracted from the perceptive observation of the crying waves recordings is then used as a help to emit diagnosis. With the idea of helping to make the whole process easier and faster we are developing a method to automatically identify, measure and highlight selected qualitative features in infant cry recordings. The processing of this identifier starts with the automatic discovery of infant cry units, which is performed by the use of a threshold applied to the energy of the signal along with another threshold applied to eliminate inspiratory cry segments, when not needed. From all the detected cry units, the process automatically identifies melodic shape, shifts, glides and noise concentration. In this work, we present, besides a quick review of related works, and a description of the perceptive analysis to help diagnosis, the process implementation, some experiments as well as the experimental results obtained.

Adult Perceptions of Pain and Hunger Cries: A Synchrony of Arousal

180 male and female nonparent adults rated tape-recordings of the initial, middle, and final 10-sec segments of pain and hunger cries on 4 7-point Likert-type scale items describing how urgent, arousing, aversive, and sick the cry segments sounded. Multivariate analyses of variance showed that the final segment of the pain cry was perceived as less urgent, arousing, and aversive than the initial and middle segments. The hunger cry was perceived as increasingly more urgent, arousing, and aversive from the initial to the middle to the final cry segments, with the final segment receiving higher ratings than the final segment of the pain cry. The middle segment of both cries was the most sick sounding. While females were more aroused than males as the hunger cry intensified, and females were less aroused than males as the pain cry subsided, the initial segments of the pain cry were particularly potent stimuli to both males and females. These results suggest that different segments of cries resulting from the same stimulus provide different messages that communicate the presumed level of infant arousal.

The Child Abuse Potential Inventory: a study of concurrent validity.

Forty-four subjects were assigned to a High or Low Abuse Potential Group based upon their Child Abuse Potential (CAP) Inventory's abuse scale scores. Each subject's heart rate (HR) and skin conductance (SC) were monitored while the subject was viewing videotape presentations of an infant smiling, crying, or quiescent. The High Abuse Potential Group had a faster peak HR throughout the videotape presentations with no significant differences among videotape segments. The Low Abuse Potential Group had a slower peak HR and had a significantly lower HR during the Smile and Cry segments than in the second Quiescent segment. The males accounted for most of the significant SC variability among tape segments. The pattern of SC responses revealed that the males were less aroused than the females, the Low Abuse Potential Group was less aroused than the High Abuse Potential group, and all four groups showed a decrease in arousal toward the end of the videotape presentations. The results provide support for the concurrent validity of the CAP inventory.

Differential parental response to familiar and unfamiliar infant distress signals

Sixteen mothers and sixteen fathers of four- to six-month-old infants listened to the sounds of their own, and an unfamiliar female infant's anger and pain cries, and two tones, as their autonomic responses were monitored. In a second series of stimulus presentations, subjects rated their own subjective feelings with respect to unpleasantness, tension-provocation and novelty of each of the six tapes. For the four infant cry segments, subjects were asked to identify the infant (own or other baby) and the kind of cry expression heard. Mothers reacted to the tapes of their own infant's cries with a brief cardiac deceleration followed by a secondary acceleration, whereas the fathers responded to their own infant's cries with decelerations. Both mothers and fathers responded to the unfamiliar infant's cries and to the tones with heart rate decelerations. The own infant's pain cry was associated with the largest electrodermal response in both parent groups, and mothers rated this cry as the most unpleasant and tension-provoking of the sounds. Mothers and fathers were highly accurate in recognizing their child's cries, but only mothers were able to identify the type of cry accurately. The results are discussed in terms of orienting and defensive response patterns and the differential caregiving experiences of mothers and fathers.

The Role of Crying Activity in Apgar Scoring

THE APGAR SCORE1is designed to determine the physical condition or viability of the newborn 60 seconds after birth, and it has been established that there is a association between low scores and neonatal mortality.1-4Two of the five segments comprising the score involve crying, and the response sought to one of them (respiratory effort) is a strong (Table 1). The second sign(reflex irritability) lists in response to a tangential foot slap. The diagnostic value of the cry alone, reported in previous studies,5,6Msuggested the potential importance of its role within the Apgar score; thus, it seemed feasible to isolate for study the two segments, respiratory and reflex, in order to determine their relative contribution to the total score. Since Apgar2has already mentioned that some of the signs of her test may be more discriminative than others, it was decided to compare each of the signs with the other and with the total score. The cry segments and the total score were also compared with an independent measure of the newborn's crying activity taken from magnetic tape recordings made in the delivery room. In this manner, the on-the-spot Apgar evaluation of the cry