Search

Speech Emotion Recognition (SER) aims to endow machines with the intelligence in perceiving latent affective components from speech. However, the existing works on deep-learning-based SER make it difficult to jointly consider time-frequency and sequential information in speech due to their structures, which may lead to deficiencies in exploring reasonable local emotional representations. In this regard, we propose a Convolutional-Recurrent Neural Network with Multiple Attention mechanisms (CRNN-MA) for SER in this paper, including paralleled Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) modules, using extracted Mel-spectrums and frame-level features respectively, in order to acquire time-frequency and sequential information simultaneously. Further, we set three strategies for the proposed CRNN-MA: A multiple self-attention layer in the CNN module on frame-level weights, a multi-dimensional attention layer as the input features of the LSTM, and a fusion layer summarizing the features of the two modules. Experimental results on three conventional SER corpora demonstrate the effectiveness of the proposed approach through using the convolutional-recurrent and multiple-attention modules, compared with other related models and existing state-of-the-art approaches.

As a common mental disorder, depression becomes a major threat to human health and may even heavily influence one’s daily life. Considering this background, it is necessary to investigate strategies for automatically detecting depression, especially through the audio modality represented by speech segments, mainly due to the efficient latent information included in speech when describing depression. However, most of the existing works focus on stacking deep networks in audio-based depression detection, which may lead to insufficient knowledge for representing depression in speech. In this regard, we propose a deep learning model based on a parallel convolutional neural network and a transformer in order to mine effective information with an acceptable complexity. The proposed approach consists of a parallel convolutional neural network (parallel-CNN) module used to focus on local knowledge, while a transformer module is employed as the other parallel stream to perceive temporal sequential information using linear attention mechanisms with kernel functions. Then, we performed experiments on two datasets of Distress Analysis Interview Corpus-Wizard of OZ (DAIC-WOZ) and Multi-modal Open Dataset for Mental-disorder Analysis (MODMA). The experimental results indicate that the proposed approach achieves a better performance compared with the state-of-the-art strategies.

During RNA elongation, the influenza A viral (IAV) RNA-dependent RNA polymerase (RdRp) residues in the active site interact with the triphosphate moiety of nucleoside triphosphate (NTP) for catalysis. The molecular mechanisms by which they control the rate and fidelity of NTP incorporation remain elusive. Here, we demonstrated through enzymology, virology and computational approaches that the R239 and K235 in the PB1 subunit of RdRp are critical to controlling the activity and fidelity of transcription. Contrary to common beliefs that high-fidelity RdRp variants exert a slower incorporation rate, we discovered a first-of-its-kind, single lysine-to-arginine mutation on K235 exhibited enhanced fidelity and activity compared with wild-type. In particular, we employed a single-turnover NTP incorporation assay for the first time on IAV RdRp to show that K235R mutant RdRp possessed a 1.9-fold increase in the transcription activity of the cognate NTP and a 4.6-fold increase in fidelity compared to wild-type. Our all-atom molecular dynamics simulations further elucidated that the higher activity is attributed to the shorter distance between K235R and the triphosphate moiety of NTP compared with wild-type. These results provide novel insights into NTP incorporation and fidelity control mechanisms, which lay the foundation for the rational design of IAV vaccine and antiviral targets.

The severe acute respiratory syndrome coronavirus 2 (SARS-COV-2) antigenic profile evolves in response to the vaccine and natural infection-derived immune pressure, resulting in immune escape and threatening public health. Exploring the possible antigenic evolutionary potentials improves public health preparedness, but it is limited by the lack of experimental assays as the sequence space is exponentially large. Here we introduce the Machine Learning-guided Antigenic Evolution Prediction (MLAEP), which combines structure modeling, multi-task learning, and genetic algorithm to model the viral fitness landscape and explore the antigenic evolution viain silicodirected evolution. As demonstrated by existing SARS-COV-2 variants, MLAEP can infer the order of variants along antigenic evolutionary trajectories, which is also strongly correlated with their sampling time. The novel mutations predicted by MLAEP are also found in immunocompromised covid patients and newly emerging variants, like XBB1.5. The predictions of MLAEP were validated by conducting in vitro neutralizing antibody binding assay, which demonstrated that the model-generated variants displayed significantly increased immune evasion ability compared with the controls. In sum, our approach enables profiling existing variants and forecasting prospective antigenic variants, thus may help guide the development of vaccines and increase preparedness against future variants. Our model is available athttps://mlaep.cbrc.kaust.edu.sa.

ObjectivesTo analyse the current curative expenditure (CCE) of NCDs in China from 2017 to 2019.DesignA cross-sectional study. Medical institutions were collected by multistage stratified random sampling from 2017 to 2019.SettingDalian, ChinaParticipants408 institutions and 8 104 233 valid items were included in the study. NCDs patients were selected according to International Classification of Diseases-10.Primary and secondary outcome measuresCCE for NCDs was measured based on the System of Health Accounts 2011. Influenced factors were analysed by linear regression. All analyses and calculations were performed by STATA V.15.0.ResultsThe CCE of NCDs was ¥14.929 billion in 2017, ¥16.377 billion in 2018 and ¥18.055 billion in 2019, which accounted for more than 65% of total expenditure spent each year. More than 60% came from public financing. The proportion of family health financing continued to decline, reaching 31.16% in 2019. The expenditures were mainly in general hospitals, above 70%. Elderly patients account for the majority. Diseases of the circulatory system, diseases of the digestive system and neoplasms were the main NCDs. Year, age, gender, length of stay, surgery, insurance and institution level affected hospitalisation expenses.ConclusionsNCDs are the main CCE of diseases in China, and their resources are not allocated reasonably. To reduce the CCE of NCDs, the government needs to optimise resource allocation and rationalise institutional flows and functions.

Speech Emotion Recognition (SER) makes it possible for machines to perceive affective information. Our previous research differed from conventional SER endeavours in that it focused on recognising unseen emotions in speech autonomously through machine learning. Such a step would enable the automatic leaning of unknown emerging emotional states. This type of learning framework, however, still relied on manual annotations to obtain multiple samples of each emotion. In order to reduce this additional workload, herein, we propose a zero-shot SER framework employing a per-emotion semantic-embedding paradigm to describe emotions in zero-shot SER, instead of using the sample-wise descriptors. Aiming to optimise the relationship between emotions, prototypes, and speech samples, this framework includes two types of learning strategies: Sample-wise learning and emotion-wise learning. These strategies apply a novel learning process to speech samples and emotions, respectively, via specifically designed semantic-embedding prototypes. We verify the utility of these approaches by performing an extensive experimental evaluation on two corpora on three aspects, namely the influence of different types of learning strategies, emotional-pair comparison, and the selections of semantic-embedding prototypes and paralinguistic features. The experimental results indicate that it is applicable to use semantic-embedding prototypes for zero-shot emotion recognition in speech, despite the influence of choosing optimal strategies and prototypes.

Background: Noncommunicable diseases (NCDs) are the leading causes of morbidity and mortality worldwide. Understanding the distribution of diseases can provide a basis for policy formulation and intervention. This study analyses the current curative expenditure (CCE) of NCDs in China to provide health policy advice.Methods: Data were collected by multi-stage stratified random sampling from 2017-2019. The medical expenses of patients with NCDs were calculated based on “System of Health Accounts 2011”(SHA 2011), Analyze from funding sources, dimensions of institutional flow, and financing scheme. Linear regression analysis was conducted by controlling factors influencing hospitalization expenses. All analyses were performed by STATA 15.0.Results: 408 institutions and 8,104,233 valid items were included in the study. The CCE of NCDs was 14.205 billion China Yuan (CNY) in 2017, 15.914 billion CNY in 2018, and 18.055 billion CNY in 2019. More than 60% came from public financing. The proportion of family health financing continued to decline, reaching 31.16% in 2019. The expenditures were mainly in general hospitals, above 70%. Elderly patients account for the majority. Diseases of the circulatory system, Diseases of the digestive system, and Neoplasms were the main NCDs. Year, age, gender, length of stay, surgery, insurance, and institution level were the factors affecting hospitalization expenses.Conclusions: NCDs are the main CCE of diseases in China, and their resources are not allocated reasonably. To reduce the CCE of NCDs, the government needs to optimize resource allocation and rationalize institutional flows and functions.

Background: Noncommunicable diseases (NCDs) are the leading causes of morbidity and mortality worldwide. Understanding the financing and distribution of diseases can provide a basis for policy formulation and intervention. This study analyses the expenditure of NCDs in China to provide health policy advice.Methods: Data were collected by multi-stage stratified random sampling from 2017-2019. The medical expenses of patients with NCDs were calculated based on “System of Health Accounts 2011”(SHA 2011), Analyze from funding sources, dimensions of institutional flow, and financing scheme. Linear regression analysis was conducted by controlling factors influencing hospitalization expenses. All analyses were performed by STATA 15.0.Results: 408 institutions and 8,104,233 valid items were included in the study. The current curative expenditure (CCE) of NCDs was 14.205 billion China Yuan (CNY) in 2017, 15.914 billion CNY in 2018, and 18.055 billion CNY in 2019. More than 60% came from public financing. The proportion of family health financing continued to decline, reaching 31.16% in 2019. The expenditures were mainly in general hospitals, above 70%. Elderly patients account for the majority. Diseases of the circulatory system, Diseases of the digestive system, and Neoplasms were the main NCDs. Year, age, gender, length of stay, surgery, insurance, and institution level were the factors affecting hospitalization expenses.Conclusions: NCDs are the main economic burden of diseases in China, and their financing and distribution are inequality. To reduce the economic burden of NCDs, the government needs to optimize resource allocation and rationalize institutional flows and functions.

Super-Resolution (SR) refers to increasing the resolution of a signal in a variety of ways, conventionally employed in the field of image enhancement. Compared with the endeavors for super-resolution in image processing, music signals require supper-resolution to improve their quality or adapt to communication in narrow-band channel, which is also regarded as bandwidth expansion. To this end, we shed light on super-resolution for music signals using the deep learning strategy of Generative Adversarial Networks (GANs). The proposed approach feeds Shot-Time Fourier Transform (STFT) features of low-band signals to the GAN, expecting to obtain their high-band information through jointly considering content and adversarial losses. Then, we carry out experiments on MUSDB18 dataset using mixtures of music sources, in order to show the performance of the proposed approach. The experimental results indicate that the proposed approach achieves better super-resolution performances compared with interpolation and some conventional deep-learning strategies.

Sound event detection (SED) refers to recognizing sound event, in which the prevailing method of SED is to employ deep neural networks at present. However, it remains unknown on the detection performance for sound events with background speech. Thus within this paper, we propose to utilize a Convolutional Recurrent Neural Network (CRNN) in detecting multi-channel audio from overlapping sources in the speech-interference condition with multiple speakers. The approach includes two modules of Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN), in which the CNN module maps acoustic features to the input of the RNN, while the RNN aims to model the time series using Bi-directional Gated Recurrent Units (Bi-GRU). Afterwards, we employ experiments on SED data with speech interference. The experimental results indicate that the proposed approach makes it robust to detect sound events when provided speech signals as the background noise, compared with conventional SED approaches.

The task of Mask-Speech Identification (MSI) aims at judging whether a chunk of speech is pronounced when the speaker is wearing a facial mask or not. Most of the existing related research focuses on investigating the influence of wearing a mask, which only adapts in some certain cases to speech analysis. Thus in order to generalise the research on MSI, we propose an MSI approach using deep networks on Low-Level Aggregation (LLA) for speech chunks. The proposed approach benefits from data augmentation on Low-Level Descriptors (LLDs), resulting in more adaptation to deep models through inputting much more samples in training without employing pre-trained knowledge. Experiments are performed on the dataset of Mask Augsburg Speech Corpus (MSC) used in the INTERSPEECH 2020 ComParE challenge, considering the influence from employing different strategies. The experimental results show effectiveness of the proposed approach compared with the ComParE challenge baselines.

Transcription errors can substantially affect metabolic processes in organisms by altering the epigenome and causing misincorporations in mRNA, which is translated into aberrant mutant proteins. Moreover, within eukaryotic genomes there are specific Transcription Error-Enriched genomic Loci (TEELs) which are transcribed by RNA polymerases with significantly higher error rates and hypothesized to have implications in cancer, aging, and diseases such as Down syndrome and Alzheimer’s. Therefore, research into transcription errors is of growing importance within the field of genetics. Nevertheless, methodological barriers limit the progress in accurately identifying transcription errors. Pro-Seq and NET-Seq can purify nascent RNA and map RNA polymerases along the genome but cannot be used to identify transcriptional mutations. Here we present background Error Model-coupled Precision nuclear run-on Circular-sequencing (EmPC-seq), a method combining a nuclear run-on assay and circular sequencing with a background error model to precisely detect nascent transcription errors and effectively discern TEELs within the genome.

We propose a novel method to learn multiscale kernels with locally penalised discriminant analysis, namely Multiscale-Kernel Locally Penalised Discriminant Analysis (MS-KLPDA). As an exemplary use-case, we apply it to recognise emotions in speech. Specifically, we employ the term of locally penalised discriminant analysis by controlling the weights of marginal sample pairs, while the method learns kernels with multiple scales. Evaluated in a series of experiments on emotional speech corpora, our proposed MS-KLPDA is able to outperform the previous research of Multiscale-Kernel Fisher Discriminant Analysis and some conventional methods in solving speech emotion recognition.

Favipiravir (T-705) has been developed as a potent anti-influenza drug and exhibited a strong inhibition effect against a broad spectrum of RNA viruses. Its active form, ribofuranosyl-triphosphate (T-705-RTP), functions as a competitive substrate for the RNA-dependent RNA polymerase (RdRp) of the influenza A virus (IAV). However, the exact inhibitory mechanisms of T-705 remain elusive and subject to a long-standing debate. Although T-705 has been proposed to inhibit transcription by acting as a chain terminator, it is also paradoxically suggested to be a mutagen towards IAV RdRp by inducing mutations due to its ambiguous base pairing of C and U. Here, we combined biochemical assay with molecular dynamics (MD) simulations to elucidate the molecular mechanism underlying the inhibitory functions exerted by T-705 in IAV RdRp. Our in vitro transcription assay illustrated that IAV RdRp could recognize T-705 as a purine analogue and incorporate it into the nascent RNA strand. Incorporating a single T-705 is incapable of inhibiting transcription as extra natural nucleotides can be progressively added. However, when two consecutive T-705 are incorporated, viral transcription is completely terminated. MD simulations reveal that the sequential appearance of two T-705 in the nascent strand destabilizes the active site and disrupts the base stacking of the nascent RNA. Altogether, our results provide a plausible explanation for the inhibitory roles of T-705 targeting IAV RdRp by integrating the computational and experimental methods. Our study also offers a comprehensive platform to investigate the inhibition effect of antivirals and a novel explanation for the designing of anti-flu drugs.

A proliferation inducing ligand (APRIL) is a member of the TNF superfamily. It shares two receptors with B-cell activating factor (BAFF), B-cell maturation antigen (BCMA), and transmembrane activator and CAML interactor (TACI). Herein, the equine APRIL was identified from equine adipose-derived stem cell (ASC), and the protein expression of APRIL and its related molecules were detected during the adipogenic differentiation of equine ASC in vitro. The equine APRIL gene was located on chromosome 11, spans 1852 base pairs (bp). Its open reading frame covers 753 bp, encoding a 250-amino acid protein with the typical TNF structure domain. During the two weeks' adipogenic differentiation of equine ASC, although the protein expression of APRIL and TACI had an insignificant change, that of BCMA increased significantly. Moreover, with the addition of recombinant protein His6-sAPRIL, a reduced differentiation of equine ASC toward adipocyte was detected. These results may provide the basis for investigating the role of APRIL in ASC adipogenic differentiation.

Features for speech emotion recognition are usually dominated by the spectral magnitude information while they ignore the use of the phase spectrum because of the difficulty of properly interpreting it. Motivated by recent successes of phase-based features for speech processing, this paper investigates the effectiveness of phase information for whispered speech emotion recognition. We select two types of phase-based features (i.e., modified group delay features and all-pole group delay features), both which have shown wide applicability to all sorts of different speech analysis and are now studied in whispered speech emotion recognition. When exploiting these features, we propose a new speech emotion recognition framework, employing outer product in combination with power and L2 normalization. The according technique encodes any variable length sequence of the phase-based features into a fixed dimension vector regardless of the length of the input sequence. The resulting representation is fed to train a classification model with a linear kernel classifier. Experimental results on the Geneva Whispered Emotion Corpus database, including normal and whispered phonation, demonstrate the effectiveness of the proposed method when compared with other modern systems. It is also shown that, combining phase information with magnitude information could significantly improve performance over the common systems solely adopting magnitude information.

In China, parents who have lost their only child and remained childless are labelled Shidu () parents. Previous research suggests high levels of psychological distress in this population, yet little is known regarding the prevalence of prolonged grief disorder (PGD) based on the new ICD-11 formulation. The present study examined prevalence rates and associated factors of prolonged grief disorder in this population. 1030 Chinese Shidu parents (381 male, 643 female) who were recruited through convenient sampling completed questionnaires assessing grief severity. Multiple linear regression models were used to examine socio-demographic, loss-related and self-reported number of chronic physical conditions associated with PGD symptoms. Results showed prevalence rate was 35.5% based on the ICD-11 PGD criteria, which was almost twice as that of Prigerson et al. (2009) criteria. Younger age of parents, being a mother, living in a rural place, lower monthly income per capital, shorter time since loss and more comorbid chronic physical conditions were related to severer PGD symptoms. The present findings revealed high rates of PGD experienced by Chinese Shidu parents and identified key risk factors which can be used for future prevention or intervention designs in this population.

Despite the widespread use of supervised learning methods for speech emotion recognition, they are severely restricted due to the lack of sufficient amount of labelled speech data for the training. Considering the wide availability of unlabelled speech data, therefore, this paper proposes semisupervised autoencoders to improve speech emotion recognition. The aim is to reap the benefit from the combination of the labelled data and unlabelled data. The proposed model extends a popular unsupervised autoencoder by carefully adjoining a supervised learning objective. We extensively evaluate the proposed model on the INTERSPEECH 2009 Emotion Challenge database and other four public databases in different scenarios. Experimental results demonstrate that the proposed model achieves state-of-the-art performance with a very small number of labelled data on the challenge task and other tasks, and significantly outperforms other alternative methods.

One of the serious obstacles to the applications of speech emotion recognition systems in real-life settings is the lack of generalization of the emotion classifiers. Many recognition systems often present a dramatic drop in performance when tested on speech data obtained from different speakers, acoustic environments, linguistic content, and domain conditions. In this letter, we propose a novel unsupervised domain adaptation model, called Universum autoencoders, to improve the performance of the systems evaluated in mismatched training and test conditions. To address the mismatch, our proposed model not only learns discriminative information from labeled data, but also learns to incorporate the prior knowledge from unlabeled data into the learning. Experimental results on the labeled Geneva Whispered Emotion Corpus database plus other three unlabeled databases demonstrate the effectiveness of the proposed method when compared to other domain adaptation methods.

The involvement of affect information in a spoken dialogue system can increase the user-friendliness and provide a more natural way for the interaction experience. This can be reached by speech emotion recognition, where the features are usually dominated by the spectral amplitude information while they ignore the use of the phase spectrum. In this chapter, we propose to use phase-based features to build up such an emotion recognition system. To exploit these features, we employ Fisher kernels. The according technique encodes the phase-based features by their deviation from a generative Gaussian mixture model. The resulting representation is fed to train a classification model with a linear kernel classifier. Experimental results on the GeWEC database including ‘normal’ and whispered phonation demonstrate the effectiveness of our method.