Vocal Characteristics Research Articles

Bimodal Emotion Recognition Based on Vocal and Facial Features

Emotion recognition is a crucial aspect of human communication, with applications in fields such as psychology, education, and healthcare. Identifying emotions accurately is challenging, as people use a variety of signals to express and perceive emotions. In this study, we address the problem of multimodal emotion recognition using both audio and video signals, to develop a robust and reliable system that can recognize emotions even when one modality is absent. To achieve this goal, we propose a novel architecture based on well-designed feature extractors for each modality and use model-level fusion based on a TFusion block to combine the information from both sources. To be more efficient in real-world scenarios, we trained our model on a compound dataset consisting of RAVDESS, RML, and eNTERFACE'05. It is then evaluated and compared to the state-of-the-art models. We find that our approach performs close to the modern solutions and can recognize emotions accurately when one of the modalities is missing. Additionally, we have developed a real-time emotion recognition application as a part of this work.

Correlation of room acoustic parameters and noise level in eating establishments

This article addresses the impact of the occupancy level, the average acoustic absorption and the so-called acoustic capacity of a space, which is proportional with the volume and inversely proportional with the reverberation time, on the behavior of talking people in an eating establishments. Four different settings were compared: two casual dining restaurants, a self-service student canteen and a small faculty club. The Lombard effect was observed in all cases. In a restaurant with an average amount of absorbing surface of 2.4 m2 or more per person, the sound pressure level increased with more than 3 dB per doubling of the number of people. Results for the student canteen show that people started to communicate less when the number of people present was so high that the absorbing surface dropped under 1.5 m2/person (80 people). The level even stopped to increase with increasing occupancy from 150 people present and beyond, corresponding with 0.8 m2 of absorbing surface per person. This is roughly consistent with an estimated value for the acoustic capacity of that space, which was 189 people (corresponding with a table occupancy of about 72%). In the latter circumstances, the background noise level, as expressed by LA,95 was as high as 69 dB. Overcoming this level for oral communication would require a not sustainable vocal effort. In the tests performed in other restaurants, the observed occupancy was below 60%, which, thanks to the higher number of absorbing surfaces in those restaurants, was well below the acoustic capacity.

Changes in vocal characteristics of male concave-eared torrent frogs (<i>Odorrana tormota</i>) in different chorus tides

Changes in vocal characteristics of male concave-eared torrent frogs (<i>Odorrana tormota</i>) in different chorus tides

GLGAN-VC: A Guided Loss-Based Generative Adversarial Network for Many-to-Many Voice Conversion.

Many-to-many voice conversion (VC) is a technique aimed at mapping speech features between multiple speakers during training and transferring the vocal characteristics of one source speaker to another target speaker, all while maintaining the content of the source speech unchanged. Existing research highlights a notable gap between the original and generated speech samples in terms of naturalness within many-to-many VC. Therefore, there is substantial room for improvement in achieving more natural-sounding speech samples for both parallel and nonparallel VC scenarios. In this study, we introduce a generative adversarial network (GAN) system with a guided loss (GLGAN-VC) designed to enhance many-to-many VC by focusing on architectural improvements and the integration of alternative loss functions. Our approach includes a pair-wise downsampling and upsampling (PDU) generator network for effective speech feature mapping (FM) in multidomain VC. In addition, we incorporate an FM loss to preserve content information and a residual connection (RC)-based discriminator network to improve learning. A guided loss (GL) function is introduced to efficiently capture differences in latent feature representations between source and target speakers, and an enhanced reconstruction loss is proposed for better contextual information preservation. We evaluate our model on various datasets, including VCC 2016, VCC 2018, VCC 2020, and an emotional speech dataset (ESD). Our results, based on both subjective and objective evaluation metrics, demonstrate that our model outperforms state-of-the-art (SOTA) many-to-many GAN-based VC models in terms of speech quality and speaker similarity in the generated speech samples.

Singer and Genre recognition using deep learning

In this paper, we try to solve the two mostly occurred problems which are genre recognition and singer identification from common pieces of music. For that, we use a system that helps us to achieve our target that is based on a music information retrieval system, in which the system can process our files and the system can generate the desired results. The system is divided into different phases and for the music recognition, first, there is a need to separate the vocal features of the singer from the music clip because, with the background music, it’s slightly difficult to analyze the vocal of the singer, then extract music features, here we work on some of the music features like tempo, RMS, song duration, frequency, dynamic range, and tonality. The RNN (recurrent neural network) has been used for training the model and testing our data set. After the training of the models, the test classifier with different unknown audio files, and the results are 91% accurate for singer identification and 90% for music genre recognition. The system will identify the singer as well as the genre of the song from the same segment of the file. For the data set, the training data used 10 different Indian and Pakistani male and female singers’ audio files in Sindhi, Hindi, and Urdu language for six types of genres like rock, pop, jazz, disco, hip-hop, and classic. The songs were sung by only one singer, the group singing and live concerts music are limitations of our system.

A new Parkinson detection system based on evolutionary fast learning networks and voice measurements

Parkinson's disease (PD) is becoming the second most neurological syndrome of the central nervous system after Alzheimer's disease. It causes diverse symptoms which include bradykinesia (slowness of movement), voice impairments, rigidity, tremor, and poor balance. PD recognition system based on voice has founded a non-invasive alternative, but involves rather complex measurements or variables. Therefore an attention is required toward new approaches for better forecasting accuracy. In this paper, an optimal fast learning network (FLN) based on genetic algorithm (GA) was established as PD diagnosis system. FLN is a double-parallel feed-forward neural network structure, and based on GA for feature reduction and hyperparameter optimisation of the FLN, it was used as a predictive model. Finally, the conducted experiments on the Parkinson data of voice recordings over ten fold cross-validation show that proposed system is less complex and also achieved better average classification results with an accuracy of 97.47%. At the same time, it is effective in automatic identification of important vocal features. Moreover, the highest average degree of improved accuracy was (2.1%) compared with other familiar wrappers including support vector machine and K-nearest neighbours in the similar conditions.

AI-POWERED APPS TO ENHANCE NOVICE NEWSREADERS’ ENGLISH PRONUNCIATION

Among the goals that Communication students, who are ESP learners, desire to achieve is to become professional newsreaders who can deliver the news in both Indonesian and English. These novice newsreaders with little to no knowledge about news delivery in English often face problems with English pronunciation. This study aims to help forty-one novice newsreaders, who are students of the Islamic Broadcasting and Communication Department, with their English pronunciation using AI-powered apps and to reveal their perceptions of the apps. The students were involved in one cycle of Classroom Action Research (CAR), where they had to make a video of them reading English news before using the apps as the pre-test and another video after the lecturer taught them how to use the apps as the post-test. The researcher distributed a questionnaire via Google Form after the post-test to complete the data. The results show that AI-powered apps can enhance novice newsreaders’ English pronunciation, especially in the aspects of Accuracy, Key Words, Chunking and Pausing, Intonation, as well as Sounds and Vocal Features; yet, Intonation becomes the lowest improved aspect. Also, the respondents believed that the apps help them with their future job as newsreaders, are easy to use, and give instant feedbacks, which are necessary for independent learning and suitable for ESP learners. The ELSA app rose to the top of the list of the respondents’ favorites. Finally, it is recommended that future researchers carry out comparable studies that focus on one or two particular English sounds evaluated across multiple cycles.

Women’s images in G.F. Handel’s operas: gender types and their transformations

Formulation of the problem. The article analyses the types of female images in G. F. Handel’s opera creative work through dramaturgical and gender aspects. Analytical attention is aimed at: 1) the typical images of heroines from the operas «Rinaldo» (Almirena), «Agrippina» (Poppaea); 2) the images that dramaturgically go beyond typicality (Armida from the opera «Rinaldo», Ginevra from the opera «Ariodante», Alcina, Agrippina, Rodelinda, Deidamia from the operas of the same name); 3) the existence of the «travestire» tradition in the composer’s creative work and in the modern space of interpretation. Analysis of recent research and publications. The study is based on the study of the vocal art of the Baroque era (Calella; Schaal; Budden; Dean); in particular in the aspect of the issue of gender (Robins; André; Degott), analysis of publications and reviews of opera performances of different years (Martin; Camuto; Thomas; Jones; Bohmer; Lang; Loney; Mudge). The purpose of the article is to analyse individual female images from the operas of G. F. Handel in the aspect of the existing typology and interpretation of the established gender types by modern performers. The scientific novelty is connected with the formulation of the parameters of gender-interpretative analysis (4 components: main affect, timbre-role, gender type and its interpretation in modern performing practice), as well as the introduction into scientific circulation of the pair of concepts travestire / vocal travestire and their approbation on the example of interpretations of female characters in G. F. Handel’s operas. The research methodology involves the modelling method (O. OganezovaGrygorenko’s idea of «timbre-role» as a constant component of the opera genre) and gender-interpretative analysis (Nikolaevska, Gigolaeva-Yurchenko). Within the latter, along with the concept of travestire, the concept of vocal travestire (essentially it is a «vocal disguise») is proposed as an author’s new vocal and stage characteristic of the countertenor’s role, based on the mismatch between the singer’s gender and the gender type of his/her voice. Results: 1) it is indicated that starting with the opera «Radamist» in 1720 and almost throughout his whole opera career, G. F. Handel almost always provided the presence of two equal female parts in the opera – soprano and contralto. At the same time, the division into prima donna and seconda donna is purely formal, while the quality of the music was always exquisite; 2) we have analysed the images and timbre-roles of the soprano on the example of Guinevere-Dalinda (opera «Ariodante»), Alcina, Armida, Agrippina, which made it possible to single out the processes of transformation of established typical roles and characters and to mark the path of their complication, which for Handel was always subordinate to the dramatic decision; 3) taking into account the fact that the role of travestire is the external plan of gender-theatrical substitution (a woman plays a man, a man plays a woman), and vocal travestire is the internal plan, an analysis of various performances of Poppea’s arias «Bel Piacere» (R. Fleming, C. Bartoli), and Almirena’s aria «Lascia ch’io pianga» from the opera «Rinaldo» (Marilyn Horn, Joyce DiDonato, M. Caballe, S. Yoncheva, P. Pettibon, Ph. Jarussky, D. Taylor, C. Bartoli) has been presented. Conclusions. In the сonclusions, it is stated that the typology of female images from G. F. Handel’s operas generally obeys the laws of the opera-seria, but G. F. Handel significantly expanded the horizons of typical characters and characters, because he focused on outstanding (and usually specific) performers. The given examples (images of Armida, Agrippina, Alcina) demonstrate the composer’s brilliant dramaturgical sense, which allowed him to make a revolution even within typical plots. According to the results of the gender-interpretative analysis, it has been established that despite the vocal and tessitura homogeneity of the presented compositions, different gender performance results were obtained, which differed from the gender of the performer. And this is no accident: feminine, masculine, androgynous features of performance are realized through genderperforming dramaturgy, combining gender characteristics of the composer’s text and individual reading of the content of the musical composition, embodied by the means of vocal performance.

Lincoln's sparrow (Melospiza lincolnii) increases singing rate in areas with chronic industrial noise

Animal communication is effective when the transmitted signal reaches the receiver and induces a behavioural response. Increases in anthropogenic noise are altering the ways animals communicate. A common adaptation shown by bird species to anthropogenic noise is song frequency shifts, but birds may also adapt their ability to communicate in noisy environments in other ways. We compared the vocal features of Lincoln's Sparrow Melospiza lincolnii breeding in noisy areas (compressor stations) and in quiet areas in Northern Alberta, Canada. We hypothesized that Lincoln's Sparrow could (1) shift its songs to avoid masking, resulting in higher‐frequency songs, and/or (2) change song allocation (singing rate, number of songs), resulting in improved song transmission. Additionally, we tested the effect of distance and height on song transmission close to a noisy compressor station and in a quiet area to determine if behavioural adaptations might allow more effective song transmission. We did not find frequency shifts between quiet and noisy areas. However, we found birds singing at higher song rates in noisy areas, supporting the song allocation hypothesis. Interestingly, the number of song types sung in noisy sites was fewer than in quiet areas. Distance influenced song transmission differently in noisy and quiet areas, as the signal to noise ratio was substantially less at 20 m in noisy areas than in the control. A higher singing rate may therefore compensate for reduced song transmission by increasing the likelihood of being heard by conspecifics in noisy areas. We found evidence of song adaptation by Lincoln's Sparrow to deal with anthropogenic noise, but whether it is sufficient to facilitate long‐distance communication used to attract females and in male–male interactions remains unknown.

Detection of Suicide Risk Using Vocal Characteristics: Systematic Review.

In an age when telehealth services are increasingly being used for forward triage, there is a need for accurate suicide risk detection. Vocal characteristics analyzed using artificial intelligence are now proving capable of detecting suicide risk with accuracies superior to traditional survey-based approaches, suggesting an efficient and economical approach to ensuring ongoing patient safety. This systematic review aimed to identify which vocal characteristics perform best at differentiating between patients with an elevated risk of suicide in comparison with other cohorts and identify the methodological specifications of the systems used to derive each feature and the accuracies of classification that result. A search of MEDLINE via Ovid, Scopus, Computers and Applied Science Complete, CADTH, Web of Science, ProQuest Dissertations and Theses A&I, Australian Policy Online, and Mednar was conducted between 1995 and 2020 and updated in 2021. The inclusion criteria were human participants with no language, age, or setting restrictions applied; randomized controlled studies, observational cohort studies, and theses; studies that used some measure of vocal quality; and individuals assessed as being at high risk of suicide compared with other individuals at lower risk using a validated measure of suicide risk. Risk of bias was assessed using the Risk of Bias in Non-randomized Studies tool. A random-effects model meta-analysis was used wherever mean measures of vocal quality were reported. The search yielded 1074 unique citations, of which 30 (2.79%) were screened via full text. A total of 21 studies involving 1734 participants met all inclusion criteria. Most studies (15/21, 71%) sourced participants via either the Vanderbilt II database of recordings (8/21, 38%) or the Silverman and Silverman perceptual study recording database (7/21, 33%). Candidate vocal characteristics that performed best at differentiating between high risk of suicide and comparison cohorts included timing patterns of speech (median accuracy 95%), power spectral density sub-bands (median accuracy 90.3%), and mel-frequency cepstral coefficients (median accuracy 80%). A random-effects meta-analysis was used to compare 22 characteristics nested within 14% (3/21) of the studies, which demonstrated significant standardized mean differences for frequencies within the first and second formants (standardized mean difference ranged between -1.07 and -2.56) and jitter values (standardized mean difference=1.47). In 43% (9/21) of the studies, risk of bias was assessed as moderate, whereas in the remaining studies (12/21, 57%), the risk of bias was assessed as high. Although several key methodological issues prevailed among the studies reviewed, there is promise in the use of vocal characteristics to detect elevations in suicide risk, particularly in novel settings such as telehealth or conversational agents. PROSPERO International Prospective Register of Systematic Reviews CRD420200167413; https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42020167413.

Age and sex differences in in-air vocalization characteristics of spotted seal pups from newborn to 1 year old in captivity

Acoustic parameters of spotted seals (Phoca largha), including the duration, peak frequency, and peak-to-peak source level, are reported to vary between different ages and sexes; however, to our knowledge, the vocal ontogeny of the spotted seal from newborn to 1 year old is yet to be studied. In the present study, we recorded and analyzed vocalizations of human-cared spotted seals from the Liaodong Bay colony from newborn to 1 year old, aiming to document the first-year development of seal vocalizations. We divided the spotted seal pups into four age groups (i.e., 1–3-month, 4–6-month, 7–9-month, and 10–12-month groups) for both sexes. The results show significant differences in sex patterns in sound parameters in terms of duration, peak frequency, and peak-to-peak source level. The vocalizations of female seal pups were longer in duration but lower in peak frequency and peak-to-peak source level than those of male pups. All three sound parameters were significantly different across different age groups. Specifically, the 1–3-month group had significantly lower values in duration, peak frequency, and peak-to-peak source level than the three other age groups. The 10–12-month group had significantly higher values in duration and peak-to-peak source level than the three other age groups. Our results also indicate a sex-specific development pattern of seal vocalizations from 1 year old until sexual maturity. Our findings will benefit the evaluation of anthropogenic noise impacts on spotted seal pups and further conservation of the seal population.

Using Vocal Characteristics To Classify Psychological Distress in Adult Helpline Callers: Retrospective Observational Study

Elevated psychological distress has demonstrated impacts on individuals' health. Reliable and efficient ways to detect distress are key to early intervention. Artificial intelligence has the potential to detect states of emotional distress in an accurate, efficient, and timely manner. The aim of this study was to automatically classify short segments of speech obtained from callers to national suicide prevention helpline services according to high versus low psychological distress and using a range of vocal characteristics in combination with machine learning approaches. A total of 120 telephone call recordings were initially converted to 16-bit pulse code modulation format. Short variable-length segments of each call were rated on psychological distress using the distress thermometer by the responding counselor and a second team of psychologists (n=6) blinded to the initial ratings. Following this, 24 vocal characteristics were initially extracted from 40-ms speech frames nested within segments within calls. After highly correlated variables were eliminated, 19 remained. Of 19 vocal characteristics, 7 were identified and validated as predictors of psychological distress using a penalized generalized additive mixed effects regression model, accounting for nonlinearity, autocorrelation, and moderation by sex. Speech frames were then grouped using k-means clustering based on the selected vocal characteristics. Finally, component-wise gradient boosting incorporating these clusters was used to classify each speech frame according to high versus low psychological distress. Classification accuracy was confirmed via leave-one-caller-out cross-validation, ensuring that speech segments from individual callers were not used in both the training and test data. The sample comprised 87 female and 33 male callers. From an initial pool of 19 characteristics, 7 vocal characteristics were identified. After grouping speech frames into 2 separate clusters (correlation with sex of caller, Cramer's V =0.02), the component-wise gradient boosting algorithm successfully classified psychological distress to a high level of accuracy, with an area under the receiver operating characteristic curve of 97.39% (95% CI 96.20-98.45) and an area under the precision-recall curve of 97.52 (95% CI 95.71-99.12). Thus, 39,282 of 41,883 (93.39%) speech frames nested within 728 of 754 segments (96.6%) were classified as exhibiting low psychological distress, and 71455 of 75503 (94.64%) speech frames nested within 382 of 423 (90.3%) segments were classified as exhibiting high psychological distress. As the probability of high psychological distress increases, male callers spoke louder, with greater vowel articulation but with greater roughness (subharmonic depth). In contrast, female callers exhibited decreased vocal clarity (entropy), greater proportion of signal noise, higher frequencies, increased breathiness (spectral slope), and increased roughness of speech with increasing psychological distress. Individual caller random effects contributed 68% to risk reduction in the classification algorithm, followed by cluster configuration (23.4%), spectral slope (4.4%), and the 50th percentile frequency (4.2%). The high level of accuracy achieved suggests possibilities for real-time detection of psychological distress in helpline settings and has potential uses in pre-emptive triage and evaluations of counseling outcomes. ANZCTR ACTRN12622000486729; https://www.anzctr.org.au/ACTRN12622000486729.aspx.

Vocal characteristics of accuracy in eyewitness testimony

In two studies, we examined if correct and incorrect testimony statements were produced with vocally distinct characteristics. Participants watched a staged crime film and were interviewed as eyewitnesses. Witness responses were recorded and then analysed along 16 vocal dimensions. Results from Study 1 showed six vocal characteristics of accuracy, which included dimensions of frequency, energy, spectral balance and temporality. Study 2 attempted to replicate Study 1, and also examined effects of emotion on the vocal characteristic-accuracy relationship. Although the results from Study 1 were not directly replicated in Study 2, a mega-analysis of the two datasets showed four distinct vocal characteristics of accuracy; correct responses were uttered with a higher pitch (F0 [M]), greater energy in the first formant region (F1 [amp]), higher speech rate (VoicedSegPerSec) and shorter pauses (UnvoicedSegM). Taken together, this study advances previous knowledge by showing that accuracy is not only indicated by what we say, but also by how we say it.

In utero exposure to cannabidiol disrupts select early-life behaviors in a sex-specific manner

Cannabidiol (CBD), one of the main components of cannabis, is generally considered safe. CBD crosses the placenta and its use during pregnancy is steadily increasing, the impact of gestational CBD’s effects on prenatal life and neurodevelopment are poorly understood. Here, we combined behavioral approaches and deep learning analysis to assess the sex-dependent neonatal behavior of CBD exposed progeny. Gestating C57BL6/J dams were exposed daily with vehicle or CBD (3 mg/Kg, s.c.), from gestational day 5 to 18. Body weight, pup ultrasound vocalizations (USVs, PND 10) and homing behavior (PND 13) were quantified in the progeny. Thus, male (but not female) pups from CBD-treated dams gained more weight than sham. There were sex-dependent differences in the coarse characteristics of ultrasonic vocalizations. Prenatally-CBD exposed male pups emitted shorter calls, whereas CBD females made more high frequency calls when compared with their control counterparts. There were significant qualitative changes in the syllabic USV repertoire reflected in call typologies and communication patterns. Finally, the homing behavior test showed that CBD-exposed females presented a greater vulnerability to gestational CBD than males. Only CBD-exposed female pups showed reduced motor and discriminatory abilities. Together the results suggest a sexual divergence in the consequences of in utero CBD exposure on neonates at early developmental ages, which may be predictive of adult psychopathology. Given the extent of cannabis and CBD use worldwide, these findings challenge the idea that CBD is a universally safe compound and reveal the need for additional studies on the effect of perinatal CBD exposure.

Vocal repertoire and sound characteristics in the variegated cardinalfish, Fowleria variegata (Pisces: Apogonidae).

The variegated cardinalfish Fowleria variegata produces grunt and hoot calls during agonistic and courtship interactions. Both sounds are tonal and occur as single and multiunit calls. Grunts are of short duration with variable frequency spectra. Hoots are longer, have a higher fundamental frequency, and a more developed harmonic structure. Agonistic grunt calls and short hoot calls (1-2 hoots) are produced during chases and when striking an individual or a mirror. Grunts are produced primarily in male-female and mirror-image encounters, and short hoot calls are produced primarily in male-male interactions. During the reproductive period, long hoot calls (three and four hoots) are the main sound type in a mix-sexed tank and at Dongsha Atoll. These are likely produced by males because isolated females are silent, and isolated males emit long hoot calls. Courtship interactions are mostly silent, and males are silent after capturing eggs for oral brooding. Tank sounds peak at dusk to early evening with a smaller peak at noon, although there are dusk and dawn peaks at Dongsha Atoll. Tank sounds exhibit a semilunar rhythm with peaks at the new and full moon. Other cardinalfish species from the atoll produce grunts but not hoot calls.

Synchronic Variation and Diachronic Evolution of Tone Patterns in Kunming Dialect

An acoustic experiment was conducted to analyze the tonal patterns of single characters in the Kunming dialect. The experimental results show that the Kunming dialect has a semi-high flat tone for the first tone, a low descending tone for the second tone, a high descending tone for the third tone, and a low tone for the fourth tone. The first and third tones are both high, while the second and fourth are both low tones, as opposed to the tonal patterns of "flat" and "low". The newer Kunming dialects show variation. The first tone starts and ends at a higher level, the third tone's bent section disappears, and the fourth tone starts and ends at a lower level, with a shorter tone duration. The variation in tone pattern starts in the middle-aged group in the old urban, and the degree of variation is female > male, old urban > old suburban, and the youth and middle-aged group > senior group, respectively. The female tone pattern has more similar vocal tone characteristics to those in the middle-aged and youth groups.

Dementia vs Depression : new methods for differential diagnosis using automatic speech analysis

AbstractBackgroundDepression is a very frequent pathology among elderly people. Depressive symptoms are common in neurocognitive disorders which can make it challenging to timely detect and diagnose them. Indeed, there is a high comorbidity of early‐stage dementia and late‐ life depression. There are only a few diagnostic tools that help support differential diagnosis. Speech analysis and the extraction of vocal and linguistic markers could be a non‐invasive new method to support this process.MethodsIn this pilot study, we included 12 female patients aged over 60 diagnosed with depression without a neurodegenerative disease or other psychiatric disorders. Clinical assessments include the MMSE, MADRS, EGF and ADC. They were asked to perform two short cognitive vocal tasks: a verbal semantic fluency (SVF) and a short narrative task (positive/negative storytelling) which were recorded. Then, in a second step, vocal and prosodic features were extracted from the audio files using machine learning methods. In a final step, we compared these features to those from a cohort of 9 matched patients with Mild Cognitive Impairment (MCI) diagnosis without any mood disorder and 21 matched controls.ResultsA non‐parametric kruskal‐wallis and BH‐adjusted post‐hoc wilcox test analysis revealed a statistically significant (p<0.05) lower word and semantic switch count for MCI compared to depression in the SVF. Linear mixed effects models were used to evaluate 18 prosodic features from the narrative speech task. Four features were significant between diagnoses; sound to noise ratio, delta deltas and deltas were significantly different for the depressed group in comparison to HC and MCI; total phonation time was longer for depressed patients than MCI patients in the negative story.ConclusionAutomatic analysis of vocal tasks such as semantic verbal fluency and storytelling could be a fast solution for supporting early characterization of patients. Certain types of extracted speech features show different patterns and namely prosodic features should be further explored. To conclude, this new method could serve as an objective assessment tool for differentiating depression from mild cognitive impairment.

Analysis of Parkinson's Disease Using an Imbalanced-Speech Dataset by Employing Decision Tree Ensemble Methods.

Parkinson's disease (PD) currently affects approximately 10 million people worldwide. The detection of PD positive subjects is vital in terms of disease prognostics, diagnostics, management and treatment. Different types of early symptoms, such as speech impairment and changes in writing, are associated with Parkinson disease. To classify potential patients of PD, many researchers used machine learning algorithms in various datasets related to this disease. In our research, we study the dataset of the PD vocal impairment feature, which is an imbalanced dataset. We propose comparative performance evaluation using various decision tree ensemble methods, with or without oversampling techniques. In addition, we compare the performance of classifiers with different sizes of ensembles and various ratios of the minority class and the majority class with oversampling and undersampling. Finally, we combine feature selection with best-performing ensemble classifiers. The result shows that AdaBoost, random forest, and decision tree developed for the RUSBoost imbalanced dataset perform well in performance metrics such as precision, recall, F1-score, area under the receiver operating characteristic curve (AUROC) and the geometric mean. Further, feature selection methods, namely lasso and information gain, were used to screen the 10 best features using the best ensemble classifiers. AdaBoost with information gain feature selection method is the best performing ensemble method with an F1-score of 0.903.

PHONETIC STATUS OF DUAL VOICES COMMON ROOT VERBS (BASED ON DLT, KYRGYZ, TURKISH)

The phonetic situations of common root verbs with two voices (VC and CV) in Kyrgyz and Turkish were examined. For the common root verbs discussed, first of all, two-voice common root verbs in Kyrgyz and Turkish were scanned. Then, it was checked whether there were verbs by comparing them with DLT (Дивану лугати’т-түрк). There is not only Kyrg.: ık- Turk.: ıh- in DLT. In the study, the consonant and vocal features of common root verbs are mentioned. Those that do not have any phonetic features are given under a separate heading on the basis of DLT only comparatively. The precise boundaries of the study were taken into account. Then, some determinations were made about the phonetic features of common root verbs in both languages. In order to determine the status of the related words, "Kâşgarlı Mahmud Dîvânu Lugâti't-Türk (2014)", "Kyrgyz Language Dictionary I-II (2011) and "Turkish Dictionary (2011)" were taken as basis.

Asymmetric Male Mating Success in Lek-Breeding Rhinella arenarum.

Mate choice is the attempt of an individual to gain higher reproductive fitness by preferring to mate with some individuals and not with others. We studied the role of mate choice in the mating system of the neotropical toad Rhinella arenarum by assessing male reproductive tactics for mate acquisition and the contribution of female choice for pair formation. In a shallow pond in central Argentina, we estimated male mating success and the corresponding reproductive tactics by focal observation. The variation of phenotypic and genotypic traits (size and shape, longevity, vocalization features, heterozygosity) was related to the observed mating success in 110 males. The phonotactic response of 21 reproductive females to conspecific advertisement call features was tested in arena experiments. Mating success was limited to 32 males, pair formation was size-assortative. The dominant reproductive tactics were advertising from call positions near suitable breeding sites and pre-mating fights with intruding males, whereas the interception of amplectant pairs and the displacement of mated males were never observed. Female phonotaxis was directed to conspecific choruses but complex and simple call structures were not distinguished. We conclude that the mating system is a lek combining pre-mating fights among males and female choice of slightly smaller males. Fights interfere with female choice, undermining size-assortative mating. This is a unique system in the R. marina species group, in which interception behavior dominates reproduction.