Sound Patterns Research Articles

Automatic detection and prediction of COVID-19 in cough audio signals using coronavirus herd immunity optimizer algorithm

The global spread of COVID-19, particularly through cough symptoms, necessitates efficient diagnostic tools. COVID-19 patients exhibit unique cough sound patterns distinguishable from other respiratory conditions. This study proposes an advanced framework to detect and predict COVID-19 using deep learning from cough audio signals. Audio data from the COUGHVID dataset undergo preprocessing through fuzzy gray level difference histogram equalization, followed by segmentation with a U-Net model. Key features are extracted via Zernike Moments (ZM) and Gray Level Co-occurrence Matrix (GLCM). The Enhanced Deep Neural Network (EDNN), tuned by the Coronavirus Herd Immunity Optimizer (CHIO), performs final prediction by minimizing error metrics. Comparative simulation results reveal that the proposed EDNN–CHIO model improves MSE by 25.35% and SMAPE by 42.06% over conventional models like PSO, WOA, and LSTM. The proposed approach demonstrates superior error reduction, highlighting its potential for effective COVID-19 detection.

Sound recurrence analysis for acoustic scene classification

In everyday life, people experience different soundscapes in which natural sounds, animal noises, and man-made sounds blend together. Although there have been several studies on the importance of recurring sound patterns in music and language, the relevance of this phenomenon in natural soundscapes is still largely unexplored. In this article, we study the repetition patterns of harmonic and transient sound events as potential cues for acoustic scene classification (ASC). In the first part of our study, our aim is to identify acoustic scene classes that exhibit characteristic sound repetition patterns concerning harmonic and transient sounds. We propose three metrics to measure the overall prevalence of sound repetitions as well as their repetition periods and temporal stability. In the second part, we evaluate three strategies to incorporate self-similarity matrices as an additional input feature to a convolutional neural network architecture for ASC. We observe the characteristic repetition of transient sounds in recordings of “park” and “street traffic” as well as harmonic sound repetitions in acoustic scene classes related to public transportation. In the ASC experiments, hybrid network architectures, which combine spectrogram features and features from sound recurrence analysis, show increased accuracy for those classes with prominent sound repetition patterns. Our findings provide additional perspective on the distinctions among acoustic scenes previously primarily ascribed in the literature to their spectral features.

Sound image as a tool of representation of difficult heritage

The article is devoted to the analysis of the sound image as a structural component of cultural memory in the context of the problem of difficult heritage. The explication of the role of sound in the construction of collective memory of traumatic experience is carried out on the example of revealing sound patterns in films representing Stalinist repression. The subject of this study will be the soundscapes captured in film language, through which the experience of collective traumatisation is revealed to us. It is proposed to understand difficult heritage as a form of cultural memory generated by trauma, which is the battleground of two opposing forces: official rhetoric that displaces the memory of traumatic experiences as a threat to collective identity, and preservationist discourse that points to the dangers of forgetting. This paper is an attempt to identify how this contradictory nature of questioning the past is represented through sound. On the basis of comparative analysis, the author identifies key features that unite the soundscape of the films under consideration. Firstly, the stratification of the auditory field into dichotomous facets: musical harmonic organisation, embodying the unifying power rhetoric, and noisy disorganisation, embodying the refusal to obscure and erase the past. Secondly, the difference in the sonic representation of the theme of repression in the cinema of the perestroika era and contemporary films, due to the change of political orientation in relation to the events of the past. The author stresses that the presence of such features relies on the existing mechanisms of labelling power instances and subordinate groups with polar sound characteristics. The author concludes that through the examination of sound recordings of the past it is possible to discover those semantic nuances for which there is no place in visual and written sources that rely on reasoned articulation.

Transformative Advancements in Fetal Cardiac Health through BILSTM Networks for FPCG Classification

This article unveils a cutting-edge methodology for Fetal Phonocardiogram(FPCG) classification employing Bidirectional Long Short-Term Memory (BILSTM)networks. Acknowledging the pivotal role of fetal heart monitoring inearly anomaly detection, the research delves into the profound insights offeredby FPCG signals concerning fetal cardiac activity. The innovative approachencompasses preprocessing FPCG signals using Mel-frequency cepstral coefficients(MFCC) and spectrogram features, coupled with the strategic applicationof BI-LSTM networks, ensuring a resilient classification framework. The bidirectionalnature of the LSTM architecture elevates the model’s ability to capturetemporal dependencies in both forward and backward directions, facilitating thediscernment of intricate patterns in fetal heart sounds. Remarkably, experimentalfindings demonstrate a remarkable 98% accuracy, reaffirming the effectivenessand precision of the BILSTM approach in fetal PCG classification. This pioneeringresearch significantly advances automated methods for evaluating fetalcardiac health, promising transformative enhancements in prenatal care.

A Speech Delay in Children: Effects of Parents' Education Level and Risk Factors

ABSTRACT Introduction : Speech delay is defined as when a child is difficult to understand compared to his or her age, or characterized by a pattern of speech error sounds that are inappropriate for children his age. Children with speech delays have a higher risk of experiencing social, emotional, behavioral, and cognitive problems in adulthood. Previous research has shown that between 2.3% and 19% of children aged 2-7 years are estimated to have speech delays. The purpose of this study was to look at the effect between parents' education level on speech delay in children and to analyze risk factors that may affect speech delay in children. Methods : The design of this study was cross sectional with a total of 95 children aged 2-7 years who came to the Medical Rehabilitation Installation, 51 children with speech delays and 44 children did not experience speech delays as controls. Results : The results of this study found that there was no statistical relationship between parents' education level and speech delays in children, both the father's education level p = 0.151 and the mother's education level p = 0.660. There was a statistically significant association between history of asphyxia (p = 0.023), comorbid presence (p = < 0.001), noise exposure (p = 0.047) and lack of stimulation (p = <0.001) on speech delay in children. Conclusion: is that there is no relationship between parents' education level and speech delays in children, influential medical factors are a history of asphyxia and the presence of comorbidities, while non-medical environmental factors that influence are noise exposure in the living environment and lack of stimulation in children.

Cognate reflex prediction as hypothesis test for a genealogical relation between the Panoan and Takanan language families.

We present a novel approach for testing genealogical relations between language families. Our method, which has previously only been applied to closely related languages, makes predictions for cognate reflexes based on the regularity of proposed sound correspondences between language families that are hypothesized to be related. We test the hypothesis about a genealogical relation between Panoan and Takanan, two linguistic families of the Amazon. The workflow contributes to new ideas of hypothesis testing in historical linguistics and can likely be transferred to other language families. We predict 206 cognate reflexes from Shipibo-Konibo, a Panoan language, from independently proposed Proto-Takanan reconstructions and test our predictions in elicitation sessions with speakers of the language. We found 21 correct predictions from the core-set, as well as another 20 correct predictions from the extended set of predictions. In addition to confirming the previously established sound correspondence patterns, we find further evidence for additional patterns that suggest the reconstruction of three new phonemes for Proto-Pano-Takanan. PROTOCOL REGISTRATION: The stage 1 protocol for this Registered Report was accepted in principle on 06/05/24. The protocol, as accepted by the journal, can be found at: https://doi.org/10.17605/OSF.IO/FGBM7 .

Stereotyped but irrelevant: frequency modulation does not promote sound pattern recognition or auditory grouping in grey treefrogs (Hyla versicolor)

ABSTRACT Acoustic signals have stereotyped features that serve numerous functions in communication. Stereotypy by itself, however, is insufficient evidence of function. In this study of the eastern grey treefrog (H. versicolor), we tested the hypotheses that a stereotyped acoustic property of advertisement calls functions in sound pattern recognition and auditory grouping. Male H. versicolor produce pulsatile calls to attract females. Each pulse comprises two harmonics that invariably sweep upwards in frequency. In a series of two-alternative phonotaxis tests, females were given choices between a standard call with a natural pattern of upward-sweeping, within-pulse frequency modulation (FM) and an alternative call in which within-pulse FM was manipulated. Manipulations included the elimination of FM and reversals of FM direction (testing the role of upward-sweeping FM in sound pattern recognition) and incoherence in the direction of FM between the two harmonics (testing the role of coherent FM in auditory grouping). Females showed no preferences for the standard call having the natural pattern of upward-sweeping FM within pulses. This result suggests within-pulse FM is an invariant call property that is irrelevant for sound pattern recognition and auditory grouping in the context of mate selection. Possible mechanistic constraints are discussed.

Musk deer are inherited

Abstract This paper explores a rare sound correspondence pattern found in Gyalrongic languages, focusing on two etyma: ‘musk deer’ and ‘valley’. The reflexes in daughter languages exhibit different onsets, including palatals, alveolars, and laterals, which renders their genetic relationship non-obvious. Despite its significance, this correspondence pattern has been largely overlooked and insufficiently reconstructed. In this study, I propose a novel reconstruction of the onset, namely *t.ɬj-, elucidating the distinct reflexes observed in different sub-branches. Through this analysis, I demonstrate that both etyma have Sino-Tibetan origins. Additionally, the paper emphasises the importance of studying correspondences with limited examples and discusses the broader implications of such research for other domains.

Cyclical variations in acoustic features within the song sessions of humpback whales (Megaptera novaeangliae).

Song sessions produced by humpback whales recorded off the coasts of Hawaii and Colombia show recurrent patterns of acoustic variation across consecutive songs. Analyses of intra-individual variations within continuous song sessions revealed that songs consistently cycled through stages of acoustic complexity. At time scales spanning tens of minutes, cyclical variations likely reflected the diving behavior of singers. Changes over shorter time frames suggest that singing humpback whales may also systematically modulate the acoustic complexity of individual sounds during song production, both by gradually morphing units and by varying the number of times they repeat sound patterns. Comparable cycles were evident across years and populations. Cyclical variations within song sessions can reveal how much time and energy singers spend producing tonal versus frequency-modulated/broadband elements. Tonal components are generally more difficult to localize spatially but easier to detect over long distances, suggesting that singing humpback whales may dynamically vary sound production in ways that affect both sound transmission and auditory spatial processing. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Voice Activated Clap Switch using Arduino with AI Technology

The development of a voice-activated clap switch system that incorporates Artificial Intelligence (AI) and Arduino technology holds significant potential for advancing smart home automation, particularly for individuals with disabilities and the elderly. In India, 4.3% of the population has disabilities, and 10.5% are over the age of 60 (Ministry of Statistics and Programme Implementation, 2021), highlighting the need for accessible technologies that facilitate independent living. Traditional clap switches, which rely solely on sound intensity detection, are prone to false triggers and background noise interference, limiting their reliability and effectiveness (Smith et al., 2020). This paper presents an AI-enhanced clap switch system capable of recognizing specific sound patterns, enabling it to accurately distinguish claps from other noises while also supporting voice commands for improved control and accessibility. The methodology involves collecting sound data, training an AI model to identify claps and process voice commands, and deploying the model on an Arduino microcontroller that controls a relay for appliance operation. Experimental results show that the AI-powered system significantly reduces false activations in noisy environments, providing a reliable, energy-efficient, and hands-free solution for smart homes, particularly benefiting users who require enhanced accessibility in their living spaces. Keywords: Voice Command, Clap Switch, AI, Smart Home Automation, Arduino.

Neural Networks for Audio Classification: Multi-scale CNN-LSTM Approach to Animal Sound Recognition

Abstract. This research explores the application of neural networks, specifically CNN-LSTM models, for classifying sound signals from dogs, frogs, and cats, selected from the ESC-50 dataset. The sound data was preprocessed using Mel-frequency cepstral coefficients (MFCCs) and augmented through time stretching, pitch shifting, and noise addition to enhance model generalization in varied acoustic environments. We compared two deep learning models: a traditional CNN-LSTM and an improved version with multi-scale feature extraction, allowing for capturing both short-term and long-term sound patterns. Our findings show that the multi-scale CNN-LSTM architecture outperforms the traditional CNN-LSTM, achieving a test accuracy of 86.11% compared to 80.56%. These results highlight the effectiveness of multi-scale feature extraction for handling complex audio signals. This research offers valuable insights into bioacoustics and has broader applications in areas such as environmental sound monitoring, wildlife preservation, and animal behavior analysis.

Impact of forest management on the communication distance of an endangered tree squirrel

AbstractLong‐distance acoustic signals mediate important social interactions between animals, and the structure of the environment can influence sound transmission to affect communication distance. Anthropogenic disturbances such as fire suppression alter forest structure and can potentially affect acoustic interactions by altering sound attenuation patterns. In the spruce‐fir and mixed‐conifer forests of the Pinaleño Mountains, Arizona, USA, that harbor endangered Mount Graham red squirrels (Tamiasciurus fremonti grahamensis), numerous historical anthropogenic disturbances have altered forest characteristics and contributed to habitat degradation and loss. In this study, we assessed how recent forest management treatments influenced the attenuation of red squirrel territorial rattle vocalizations. In June 2023–August 2023, we broadcast and recorded rattles 5.4 m above the forest floor at various distances (1, 10, 20, and 40 m) to mimic hypothetical senders and receivers. We used on‐the‐ground measurements and lidar to quantify forest structure relative to patterns of sound attenuation in 3 treatments: thinning, understory fuel reduction, and untreated control plots. Across all treatments, we found that increasing tree basal area resulted in higher attenuation, with rattles being approximately 6 dB lower in amplitude in the most dense compared to the least dense plots, equivalent to a doubling in squirrel communication distance. Additionally, rattles on untreated control plots experienced more attenuation than thinned (~3 dB) and fuel‐reduced (~1.5 dB) plots. A lidar‐derived measure of canopy structure was a better predictor of rattle attenuation than lower resolution on‐the‐ground measurements. In the short‐term, forest thinning may facilitate development of acoustic social neighborhoods that increase squirrel fitness, though alternative costs of more open forests exist. More generally, our results indicate that forest management can affect the efficacy of acoustic communication and that integration of signaling and sensory ecology with remote sensing can inform wildlife conservation.

Felt Anticipation: Rhythm, Expectation, and Emotion in Friederike Mayröcker’s Poetic Spoken Atmospheres

Abstract This paper addresses the work of the Austrian poet Friederike Mayröcker (1924–2021) with an interdisciplinary approach contextualised within Cognitive Poetics and Cognitive Literary Theory studies. The specific focus of this study is the voice and the meaning produced through the proposed notion of a ‘poetic spoken atmosphere’. The poetic spoken atmosphere is understood here not as the unique property of an audio-text but rather as an intersubjective phenomenon emerging at the intersection of the poetic language and a listener. This notion underlines the importance of voice and its effects in the dynamics of poetry reception in order to investigate sound patterns and meaning-making dynamics, approaching three specific processes – rhythm, expectation, and emotion. The general theoretical frame of this study is based on principles of both phenomenology and cognitive sciences. Likewise, the analysis of Mayröcker’s poems is based on her recordings, available on the Lyrikline platform of the Haus für Poesie Berlin, from which spectrograms were generated to complement the textual and acoustic analysis. The observations of this study address the vocal structures modulating what can be termed an ‘insideness’, which allows to analyze sound attributes such as duration, intensity, ubiquity, and disappearing. Recognising such attributes help us to understand the emotional level of poetic sounds, i. e. the relationship between feeling and meaning in listening, hypothesising that meaning is indeed felt. Finally, the paper offers a comparative perspective that suggests further research on the basis of the paper’s theoretical-analytical methodology, and it provides specific examples from several literary systems.

Sounds as taxonomic indicators in Holocentrid fishes.

The species-specific character of sounds in the animal kingdom has been extensively documented, yet research on fishes has predominantly focused on a limited number of species, overlooking the potential of acoustic signals to reflect broader taxonomic ranks. In this study, we analyzed acoustic data of hand-held sounds from 388 specimens spanning 5 genera and 33 species within the family Holocentridae, with the objective of evaluating the use of sound characteristics for taxonomic discrimination across various levels (subfamily, genus, species). Sounds could be indicative of grouping. Taxa discriminability depends on taxonomic level; the higher the taxonomic level, the better the discrimination of taxa based on sounds. Analogous to the role of morphological traits in taxonomic delineation, this research corroborates the utility of acoustic features in identifying fish taxa across multiple hierarchical levels. Remarkably, certain holocentrid species have evolved complex sound patterns characterized by unique temporal arrangements where pulses are not continuous but emitted in blocks, facilitating the exploitation of the acoustic space.

Speech sound errors in adolescents with congenital hearing loss

Children with hearing impairment (HI) often have vowel and consonant errors. Speech-language pathologists need this information to correct articulation errors. Despite the prevalence of research on speech sound error patterns in hearing aid users, there is a lack of literature in the adolescent group. The present study aimed to investigate the common speech sound errors in HI adolescents from the Malayalam-speaking community who have severe to profound hearing loss and are digital hearing aid users. A total of 15 monolingual Malayalam-speaking children between the ages of 9 and 16 years were recruited who had been diagnosed with speech sound disorders secondary to congenital hearing impairment. The minimum language age for inclusion in the study was over three years. Articulatory error analysis was performed for all consonants and vowels of the Malayalam language, and frequently produced vowel and consonant errors were included in the SODA and PMV analyses. The consonants frequently produced speech sound errors and no vowel errors were found. The most commonly produced consonant errors were fricatives, trills, laterals, velars, and alveolars. Substitution and omission errors were the most common type of error. Place and manner errors were more common than other types of errors. The results of this study may help parents and rehabilitation professionals to understand the common errors that persist even after intensive intervention. It may also be a useful guide for AAC therapists and speech-language pathologists in setting appropriate speech production goals for adolescents with HI.

A novel monitoring method based on the fusion of sound and image signals for laser welding penetration status

Accurately identifying penetration status is crucial for ensuring high-quality welding. However, relying on a single sensor to monitor the welding status falls short in accuracy. Thus, a novel monitoring method that fuses sound and image is proposed for assessing the laser welding penetration status. First, a microphone and a high-speed camera are employed to collect sound and image signals during welding. Second, to overcome noise challenges in sound signal, a denoising method is proposed that utilizes Grey Wolf Optimization (GWO) to find the optimal parameters for Variational Mode Decomposition (VMD). The time and frequency domain features from the denoised sound and Local Binary Pattern (LBP) features from the image are extracted. These features constitute a high-dimensional dataset, which is reduced using Principal Component Analysis (PCA). Finally, a back propagation neural network (BPNN) is constructed for fusion monitoring. The results demonstrate that fusion monitoring achieves higher accuracy (97.9%) compared to using sound (79.9%) or image (92.3%) features independently. In comparison to other methods like K-Nearest Neighbor (KNN), Naive Bayesian (NB), Support Vector Machine (SVM), and Decision Tree, BPNN fusion monitoring achieves higher accuracies, with increases of 18.09%, 12.79%, 6.53%, and 25.84% respectively. The proposed method helps to promote the development of welding intelligence.

Pleasantness makes a good time: musical consonance shapes interpersonal synchronization in dyadic joint action.

Music making is a process by which humans across cultures come together to create patterns of sounds that are aesthetically pleasing. What remains unclear is how this aesthetic outcome affects the sensorimotor interaction between participants. Here we approach this question using an interpersonal sensorimotor synchronization paradigm to test whether the quality of a jointly created chord (consonant vs. dissonant) affects movement coordination. We recruited non-musician participants in dyads to perform a dyadic synchronization-continuation task (dSCT): on each trial, participants first synchronized their movements to a metronome (synchronization phase) and then continued tapping together at the same tempo without the metronome (continuation phase). Each tap yielded a note and participants heard both their own and that of their partner, thus creating a chord that was varied to be either consonant (Perf5 or Maj6) or dissonant (Min2 or Maj2). For each trial, participants also rated the pleasure they felt in creating the sounds together. Additionally, they completed questionnaires about social closeness to the other participant, musical reward sensitivity and musical training. Results showed that participants' taps were closer in time when they jointly created consonant (high pleasure) vs. dissonant (low pleasure) chords, and that pleasure experienced by the dyad in each trial predicted interpersonal synchronization. However, consonance did not affect individual synchronization with the metronome or individual tapping when the metronome was discontinued. The effect of consonance on synchronization was greater in dyads who reported feeling less close prior to the task. Together, these results highlight the role of consonance in shaping the temporal coordination of our actions with others. More broadly, this work shows that the aesthetic outcome of what we create together affects joint behaviors.

Input, Universals, and Transfer in Developing Rhotics: A Sketch in Bilingualism

Understanding the role of input in bilingual phonological acquisition is revealing for deciphering the workings of language acquisition processes. Input and usage distributional frequencies guide and differentiate speech sound acquisition patterns cross-linguistically. Such processes are operant in first- and second-language acquisition. There is an under-representation of investigations on how context-specific input in bilingualism influences the early acquisition of rhotics in child developmental speech longitudinally. This study addresses the gap by tracing a Greek/English bilingual girl’s rhotic development between ages 2;7 and 3;11, utilizing naturalistic data during daily interactions with an adult interlocutor. The study reports and schematically illustrates the child’s bilingual usage frequencies, informing language choice in her production variables, which demonstrate, quantitatively and qualitatively, the effects of context-specific input on rhotic accuracy levels and substitution patterns in both languages. Specifically, distributional frequencies in the input govern the child’s phonemic and phonetic tendencies in the languages. Findings are compared with previous reports in the literature and enhance language acquisition theory, revealing the pivotal role of input in the dynamic interplay with developmental universals, language-specific tendencies, transfer, and individual variation.

Enhancing spelling skills in dyslexic primary students: A comparative study of remedial and compensatory interventions in chennai

Learning to spell poses significant challenges for young learners with dyslexia, who often encounter difficulties with phonological awareness and orthographic processing. This study explores the effectiveness of a structured, two-stage intervention aimed at enhancing spelling skills among 25 primary school students in Chennai, Tamil Nadu, whose native language is Tamil. In the initial four-week phase, participants engaged in remedial activities focused on developing phonological and orthographic awareness, essential foundations for accurate spelling. This phase targeted the students’ ability to recognize sound patterns and understand letter-sound associations, providing a base for further learning. In the following sixteen-week phase, the intervention expanded to include compensatory strategies designed to reinforce initial learning and foster a more comprehensive understanding of spelling. This phase incorporated the multimodal learning, combining visual, auditory, and kinesthetic techniques, as well as task-oriented exercises, play-based activities, and gamified methods to make the learning process more engaging and enjoyable. These varied approaches aimed to enhance students’ spelling skills while maintaining their interest and motivation. The study’s findings demonstrated that both the remedial and compensatory interventions had a positive impact on students' spelling abilities. However, while the phonological and orthographic methods helped build essential skills, they required a longer duration to ensure retention beyond the intervention period. Conversely, the integrated, multimodal approach introduced in the second phase resulted in faster and more substantial improvements. This outcome highlights the benefits of a dynamic, interactive approach in supporting dyslexic learners who are acquiring English as a second language, promoting both skill acquisition and engagement.

Acoustic and Clinical Data Analysis of Vocal Recordings: Pandemic Insights and Lessons.

Background/Objectives: The interest in processing human speech and other human-generated audio signals as a diagnostic tool has increased due to the COVID-19 pandemic. The project OSCAR (vOice Screening of CoronA viRus) aimed to develop an algorithm to screen for COVID-19 using a dataset of Portuguese participants with voice recordings and clinical data. Methods: This cross-sectional study aimed to characterise the pattern of sounds produced by the vocal apparatus in patients with SARS-CoV-2 infection documented by a positive RT-PCR test, and to develop and validate a screening algorithm. In Phase II, the algorithm developed in Phase I was tested in a real-world setting. Results: In Phase I, after filtering, the training group consisted of 166 subjects who were effectively available to train the classification model (34.3% SARS-CoV-2 positive/65.7% SARS-CoV-2 negative). Phase II enrolled 58 participants (69.0% SARS-CoV-2 positive/31.0% SARS-CoV-2 negative). The final model achieved a sensitivity of 85%, a specificity of 88.9%, and an F1-score of 84.7%, suggesting voice screening algorithms as an attractive strategy for COVID-19 diagnosis. Conclusions: Our findings highlight the potential of a voice-based detection strategy as an alternative method for respiratory tract screening.