Search

Audio-based classification techniques for body sounds have long been studied to aid in the diagnosis of respiratory diseases. While most research is centered on the use of coughs as the main acoustic biomarker, other body sounds also have the potential to detect respiratory diseases. Recent studies on the coronavirus disease 2019 (COVID-19) have suggested that breath and speech sounds, in addition to cough, correlate with the disease. Our study proposes fused audio instance and representation (FAIR) as a method for respiratory disease detection. FAIR relies on constructing a joint feature vector from various body sounds represented in waveform and spectrogram form. We conduct experiments on the use case of COVID-19 detection by combining waveform and spectrogram representation of body sounds. Our findings show that the use of self-attention to combine extracted features from cough, breath, and speech sounds leads to the best performance with an area under the receiver operating characteristic curve (AUC) score of 0.8658, a sensitivity of 0.8057, and a specificity of 0.7958. Compared to models trained solely on spectrograms or waveforms, the use of both representations results in an improved AUC score, demonstrating that combining spectrogram and waveform representation helps to enrich the extracted features and outperforms the models that use only one representation. While this study focuses on COVID-19, FAIR’s flexibility allows it to combine various multi-modal and multi-instance features in many other diagnostic applications, potentially leading to more accurate diagnoses across a wider range of diseases.

Radar sensors, leveraging the Doppler effect, enable the nonintrusive capture of kinetic and physiological motions while preserving privacy. Deep learning (DL) facilitates radar sensing for healthcare applications such as gait recognition and vital-sign measurement. However, band-dependent patterns, indicating variations in patterns and power scales associated with frequencies in time–frequency representation (TFR), challenge radar sensing applications using DL. Frequency-dependent characteristics and features with lower power scales may be overlooked during representation learning. This paper proposes an Enhanced Band-Dependent Learning framework (E-BDL) comprising an adaptive sub-band filtering module, a representation learning module, and a sub-view contrastive module to fully detect band-dependent features in sub-frequency bands and leverage them for classification. Experimental validation is conducted on two radar datasets, including gait abnormality recognition for Alzheimer’s disease (AD) and AD-related dementia (ADRD) risk evaluation and vital-sign monitoring for hemodynamics scenario classification. For hemodynamics scenario classification, E-BDL-ResNet achieves competitive performance in overall accuracy and class-wise evaluations compared to recent methods. For ADRD risk evaluation, the results demonstrate E-BDL-ResNet’s superior performance across all candidate models, highlighting its potential as a clinical tool. E-BDL effectively detects salient sub-bands in TFRs, enhancing representation learning and improving the performance and interpretability of DL-based models.

In this paper, we present a deep learning approach for identifying current intensity and frequency. The reconstruction is based on measurements of the magnetic field generated by the current flowing in a conductor. Magnetic field data are collected using a magnetic probe capable of generating a spectrogram, representing the spectrum of frequencies of the magnetic field over time. These spectrograms are saved as images characterized by color density proportional to the induction field value at a given frequency. The proposed deep learning approach utilizes a convolutional neural network (CNN) with the spectrogram image as input and the current or frequency value as output. One advantage of this approach is that current estimation is achieved contactless, using a simple magnetic field probe positioned close to the conductor.

Pneumonia causes the deaths of over a million people worldwide each year, with most occurring in countries with limited access to expensive but effective diagnostic methods, e.g., chest X-rays. Physical examination, the other major established method of diagnosis, suffers from several drawbacks, most notably low accuracy and high interobserver error. We sought to address this diagnostic gap by developing a proof-of-concept non-invasive device to identify the accumulation of fluid in the lungs (consolidation) characteristic of pneumonia. This device, named Tabla after the percussive instrument of the same name, utilizes the technique of auscultatory percussion; a percussive input sound is sent through the chest and recorded with a digital stethoscope for analysis. Tabla analyzes differences in sound transmission through the chest at audible frequencies as a marker for lung consolidation. This paper presents preliminary data from five pneumonia patients and eight healthy subjects. We demonstrate 92.3% accuracy in distinguishing between healthy subjects and patients with pneumonia after data analysis with a K-nearest neighbors algorithm. This prototype device is low cost and simple to implement and may offer a rapid and inexpensive method for pneumonia diagnosis appropriate for general use and in areas with limited medical infrastructure.

Understanding and identifying emotional cues in human speech is a crucial aspect of human–computer communication. The application of computer technology in dissecting and deciphering emotions, along with the extraction of relevant emotional characteristics from speech, forms a significant part of this process. The objective of this study was to architect an innovative framework for speech emotion recognition predicated on spectrograms and semantic feature transcribers, aiming to bolster performance precision by acknowledging the conspicuous inadequacies in extant methodologies and rectifying them. To procure invaluable attributes for speech detection, this investigation leveraged two divergent strategies. Primarily, a wholly convolutional neural network model was engaged to transcribe speech spectrograms. Subsequently, a cutting-edge Mel-frequency cepstral coefficient feature abstraction approach was adopted and integrated with Speech2Vec for semantic feature encoding. These dual forms of attributes underwent individual processing before they were channeled into a long short-term memory network and a comprehensive connected layer for supplementary representation. By doing so, we aimed to bolster the sophistication and efficacy of our speech emotion detection model, thereby enhancing its potential to accurately recognize and interpret emotion from human speech. The proposed mechanism underwent a rigorous evaluation process employing two distinct databases: RAVDESS and EMO-DB. The outcome displayed a predominant performance when juxtaposed with established models, registering an impressive accuracy of 94.8% on the RAVDESS dataset and a commendable 94.0% on the EMO-DB dataset. This superior performance underscores the efficacy of our innovative system in the realm of speech emotion recognition, as it outperforms current frameworks in accuracy metrics.

With the evolution in technology, communication based on the voice has gained importance in applications such as online conferencing, online meetings, voice-over internet protocol (VoIP), etc. Limiting factors such as environmental noise, encoding and decoding of the speech signal, and limitations of technology may degrade the quality of the speech signal. Therefore, there is a requirement for continuous quality assessment of the speech signal. Speech quality assessment (SQA) enables the system to automatically tune network parameters to improve speech quality. Furthermore, there are many speech transmitters and receivers that are used for voice processing including mobile devices and high-performance computers that can benefit from SQA. SQA plays a significant role in the evaluation of speech-processing systems. Non-intrusive speech quality assessment (NI-SQA) is a challenging task due to the unavailability of pristine speech signals in real-world scenarios. The success of NI-SQA techniques highly relies on the features used to assess speech quality. Various NI-SQA methods are available that extract features from speech signals in different domains, but they do not take into account the natural structure of the speech signals for assessment of speech quality. This work proposes a method for NI-SQA based on the natural structure of the speech signals that are approximated using the natural spectrogram statistical (NSS) properties derived from the speech signal spectrogram. The pristine version of the speech signal follows a structured natural pattern that is disrupted when distortion is introduced in the speech signal. The deviation of NSS properties between the pristine and distorted speech signals is utilized to predict speech quality. The proposed methodology shows better performance in comparison to state-of-the-art NI-SQA methods on the Centre for Speech Technology Voice Cloning Toolkit corpus (VCTK-Corpus) with a Spearman’s rank-ordered correlation constant (SRC) of 0.902, Pearson correlation constant (PCC) of 0.960, and root mean squared error (RMSE) of 0.206. Conversely, on the NOIZEUS-960 database, the proposed methodology shows an SRC of 0.958, PCC of 0.960, and RMSE of 0.114.

The convolution module in Conformer is capable of providing translationally invariant convolution in time and space. This is often used in Mandarin recognition tasks to address the diversity of speech signals by treating the time-frequency maps of speech signals as images. However, convolutional networks are more effective in local feature modeling, while dialect recognition tasks require the extraction of a long sequence of contextual information features; therefore, the SE-Conformer-TCN is proposed in this paper. By embedding the squeeze-excitation block into the Conformer, the interdependence between the features of channels can be explicitly modeled to enhance the model’s ability to select interrelated channels, thus increasing the weight of effective speech spectrogram features and decreasing the weight of ineffective or less effective feature maps. The multi-head self-attention and temporal convolutional network is built in parallel, in which the dilated causal convolutions module can cover the input time series by increasing the expansion factor and convolutional kernel to capture the location information implied between the sequences and enhance the model’s access to location information. Experiments on four public datasets demonstrate that the proposed model has a higher performance for the recognition of Mandarin with an accent, and the sentence error rate is reduced by 2.1% compared to the Conformer, with only 4.9% character error rate.

Bearing defects are a common problem in rotating machines and equipment that can lead to unexpected downtime, costly repairs, and even safety hazards. Diagnosing bearing defects is crucial for preventative maintenance, and deep learning models have shown promising results in this field. On the other hand, the high complexity of these models can lead to high computational and data processing costs, making their practical implementation challenging. Recent studies have focused on optimizing these models by reducing their size and complexity, but these methods often compromise classification performance. This paper proposes a new approach that reduces the dimensionality of input data and optimizes the model structure simultaneously. A much lower input data dimension than that of existing deep learning models was achieved by downsampling the vibration sensor signals used for bearing defect diagnosis and constructing spectrograms. This paper introduces a lite convolutional neural network (CNN) model with fixed feature map dimensions that achieve high classification accuracy with low-dimensional input data. The vibration sensor signals used for bearing defect diagnosis were first downsampled to reduce the dimensionality of the input data. Next, spectrograms were constructed using the signals of the minimum interval. Experiments were conducted using the vibration sensor signals from the Case Western Reserve University (CWRU) dataset. The experimental results show that the proposed method could be highly efficient in terms of computation while maintaining outstanding classification performance. The results show that the proposed method outperformed a state-of-the-art model for bearing defect diagnosis under different conditions. This approach is not limited to the field of bearing failure diagnosis, but could be applied potentially to other fields that require the analysis of high-dimensional time series data.

Recently, human–machine interfaces (HMI) that make life convenient have been studied in many fields. In particular, a hand gesture recognition (HGR) system, which can be implemented as a wearable system, has the advantage that users can easily and intuitively control the device. Among the various sensors used in the HGR system, the surface electromyography (sEMG) sensor is independent of the acquisition environment, easy to wear, and requires a small amount of data. Focusing on these advantages, previous sEMG-based HGR systems used several sensors or complex deep-learning algorithms to achieve high classification accuracy. However, systems that use multiple sensors are bulky, and embedded platforms with complex deep-learning algorithms are difficult to implement. To overcome these limitations, we propose an HGR system using a binarized neural network (BNN), a lightweight convolutional neural network (CNN), with one dry-type sEMG sensor, which is implemented on a field-programmable gate array (FPGA). The proposed HGR system classifies nine dynamic gestures that can be useful in real life rather than static gestures that can be classified relatively easily. Raw sEMG data collected from a dynamic gesture are converted into a spectrogram with information in the time-frequency domain and transferred to the classifier. As a result, the proposed HGR system achieved 95.4% classification accuracy, with a computation time of 14.1 ms and a power consumption of 91.81 mW.

Despite the lack of findings in laryngeal endoscopy, it is common for patients to undergo vocal problems after thyroid surgery. This study aimed to predict the recovery of the patient’s voice after 3 months from preoperative and postoperative voice spectrograms. We retrospectively collected voice and the GRBAS score from 114 patients undergoing surgery with thyroid cancer. The data for each patient were taken from three points in time: preoperative, and 2 weeks and 3 months postoperative. Using the pretrained model to predict GRBAS as the backbone, the preoperative and 2-weeks-postoperative voice spectrogram were trained for the EfficientNet architecture deep-learning model with long short-term memory (LSTM) to predict the voice at 3 months postoperation. The correlation analysis of the predicted results for the grade, breathiness, and asthenia scores were 0.741, 0.766, and 0.433, respectively. Based on the scaled prediction results, the area under the receiver operating characteristic curve for the binarized grade, breathiness, and asthenia were 0.894, 0.918, and 0.735, respectively. In the follow-up test results for 12 patients after 6 months, the average of the AUC values for the five scores was 0.822. This study showed the feasibility of predicting vocal recovery after 3 months using the spectrogram. We expect this model could be used to relieve patients’ psychological anxiety and encourage them to actively participate in speech rehabilitation.

Infectious diseases remain a common problem in low- and middle-income countries, including in Vietnam. Tetanus is a severe infectious disease characterized by muscle spasms and complicated by autonomic nervous system dysfunction in severe cases. Patients require careful monitoring using electrocardiograms (ECGs) to detect deterioration and the onset of autonomic nervous system dysfunction as early as possible. Machine learning analysis of ECG has been shown of extra value in predicting tetanus severity, however any additional ECG signal analysis places a high demand on time-limited hospital staff and requires specialist equipment. Therefore, we present a novel approach to tetanus monitoring from low-cost wearable sensors combined with a deep-learning-based automatic severity detection. This approach can automatically triage tetanus patients and reduce the burden on hospital staff. In this study, we propose a two-dimensional (2D) convolutional neural network with a channel-wise attention mechanism for the binary classification of ECG signals. According to the Ablett classification of tetanus severity, we define grades 1 and 2 as mild tetanus and grades 3 and 4 as severe tetanus. The one-dimensional ECG time series signals are transformed into 2D spectrograms. The 2D attention-based network is designed to extract the features from the input spectrograms. Experiments demonstrate a promising performance for the proposed method in tetanus classification with an F1 score of 0.79 ± 0.03, precision of 0.78 ± 0.08, recall of 0.82 ± 0.05, specificity of 0.85 ± 0.08, accuracy of 0.84 ± 0.04 and AUC of 0.84 ± 0.03.

Power fault monitoring based on acoustic waves has gained a great deal of attention in industry. Existing methods for fault diagnosis typically collect sound signals on site and transmit them to a back-end server for analysis, which may fail to provide a real-time response due to transmission packet loss and latency. However, the limited computing power of edge devices and the existing methods for feature extraction pose a significant challenge to performing diagnosis on the edge. In this paper, we propose a fast Lightweight Fault Diagnosis method for power transformers, referred to as LightFD, which integrates several technical components. Firstly, before feature extraction, we design an asymmetric Hamming-cosine window function to reduce signal spectrum leakage and ensure data integrity. Secondly, we design a multidimensional spatio-temporal feature extraction method to extract acoustic features. Finally, we design a parallel dual-layer, dual-channel lightweight neural network to realize the classification of different fault types on edge devices with limited computing power. Extensive simulation and experimental results show that the diagnostic precision and recall of LightFD reach 94.64% and 95.33%, which represent an improvement of 4% and 1.6% over the traditional SVM method, respectively.

When recording seismic ground motion in multiple sites using independent recording stations one needs to recognize the presence of the same parts of seismic waves arriving at these stations. This problem is known in seismology as seismic phase picking. It is challenging to automate the accurate picking of seismic phases to the level of human capabilities. By solving this problem, it would be possible to automate routine processing in real time on any local network. A new machine learning approach was developed to classify seismic phases from local earthquakes. The resulting model is based on spectrograms and utilizes the transformer architecture with a self-attention mechanism and without any convolution blocks. The model is general for various local networks and has only 57 k learning parameters. To assess the generalization property, two new datasets were developed, containing local earthquake data collected from two different regions using a wide variety of seismic instruments. The data were not involved in the training process for any model to estimate the generalization property. The new model exhibits the best classification and computation performance results on its pre-trained weights compared with baseline models from related work. The model code is available online and is ready for day-to-day real-time processing on conventional seismic equipment without graphics processing units.

Audio signal classification finds various applications in detecting and monitoring health conditions in healthcare. Convolutional neural networks (CNN) have produced state-of-the-art results in image classification and are being increasingly used in other tasks, including signal classification. However, audio signal classification using CNN presents various challenges. In image classification tasks, raw images of equal dimensions can be used as a direct input to CNN. Raw time-domain signals, on the other hand, can be of varying dimensions. In addition, the temporal signal often has to be transformed to frequency-domain to reveal unique spectral characteristics, therefore requiring signal transformation. In this work, we overview and benchmark various audio signal representation techniques for classification using CNN, including approaches that deal with signals of different lengths and combine multiple representations to improve the classification accuracy. Hence, this work surfaces important empirical evidence that may guide future works deploying CNN for audio signal classification purposes.

In this paper, we propose a method of identifying human motions, such as standing, walking, running, and crawling, using a millimeter wave radar sensor. In our method, two signal processing is performed in parallel to identify the human motions. First, the moment at which a person’s motion changes is determined based on the statistical characteristics of the radar signal. Second, a deep learning-based classification algorithm is applied to determine what actions a person is taking. In each of the two signal processing, radar spectrograms containing the characteristics of the distance change over time are used as input. Finally, we evaluate the performance of the proposed method with radar sensor data acquired in an indoor environment. The proposed method can find the moment when the motion changes with an error rate of 3%, and also can classify the action that a person is taking with more than 95% accuracy.

The appearance of wheezing sounds is widely considered by physicians as a key indicator to detect early pulmonary disorders or even the severity associated with respiratory diseases, as occurs in the case of asthma and chronic obstructive pulmonary disease. From a physician’s point of view, monophonic and polyphonic wheezing classification is still a challenging topic in biomedical signal processing since both types of wheezes are sinusoidal in nature. Unlike most of the classification algorithms in which interference caused by normal respiratory sounds is not addressed in depth, our first contribution proposes a novel Constrained Low-Rank Non-negative Matrix Factorization (CL-RNMF) approach, never applied to classification of wheezing as far as the authors’ knowledge, which incorporates several constraints (sparseness and smoothness) and a low-rank configuration to extract the wheezing spectral content, minimizing the acoustic interference from normal respiratory sounds. The second contribution automatically analyzes the harmonic structure of the energy distribution associated with the estimated wheezing spectrogram to classify the type of wheezing. Experimental results report that: (i) the proposed method outperforms the most recent and relevant state-of-the-art wheezing classification method by approximately 8% in accuracy; (ii) unlike state-of-the-art methods based on classifiers, the proposed method uses an unsupervised approach that does not require any training.

The factors affecting the penetration of certain diseases such as COVID-19 in society are still unknown. Internet of Things (IoT) technologies can play a crucial role during the time of crisis and they can provide a more holistic view of the reasons that govern the outbreak of a contagious disease. The understanding of COVID-19 will be enriched by the analysis of data related to the phenomena, and this data can be collected using IoT sensors. In this paper, we show an integrated solution based on IoT technologies that can serve as opportunistic health data acquisition agents for combating the pandemic of COVID-19, named CIoTVID. The platform is composed of four layers—data acquisition, data aggregation, machine intelligence and services, within the solution. To demonstrate its validity, the solution has been tested with a use case based on creating a classifier of medical conditions using real data of voice, performing successfully. The layer of data aggregation is particularly relevant in this kind of solution as the data coming from medical devices has a very different nature to that coming from electronic sensors. Due to the adaptability of the platform to heterogeneous data and volumes of data; individuals, policymakers, and clinics could benefit from it to fight the propagation of the pandemic.

With the upsurge in the use of Unmanned Aerial Vehicles (UAVs) in various fields, detecting and identifying them in real-time are becoming important topics. However, the identification of UAVs is difficult due to their characteristics such as low altitude, slow speed, and small radar cross-section (LSS). With the existing deterministic approach, the algorithm becomes complex and requires a large number of computations, making it unsuitable for real-time systems. Hence, effective alternatives enabling real-time identification of these new threats are needed. Deep learning-based classification models learn features from data by themselves and have shown outstanding performance in computer vision tasks. In this paper, we propose a deep learning-based classification model that learns the micro-Doppler signatures (MDS) of targets represented on radar spectrogram images. To enable this, first, we recorded five LSS targets (three types of UAVs and two different types of human activities) with a frequency modulated continuous wave (FMCW) radar in various scenarios. Then, we converted signals into spectrograms in the form of images by Short time Fourier transform (STFT). After the data refinement and augmentation, we made our own radar spectrogram dataset. Secondly, we analyzed characteristics of the radar spectrogram dataset with the ResNet-18 model and designed the ResNet-SP model with less computation, higher accuracy and stability based on the ResNet-18 model. The results show that the proposed ResNet-SP has a training time of 242 s and an accuracy of 83.39%, which is superior to the ResNet-18 that takes 640 s for training with an accuracy of 79.88%.

Motor current signature analysis (MCSA) is a fault diagnosis method for induction machines (IMs) that has attracted wide industrial interest in recent years. It is based on the detection of the characteristic fault signatures that arise in the current spectrum of a faulty induction machine. Unfortunately, the MCSA method in its basic formulation can only be applied in steady state functioning. Nevertheless, every day increases the importance of inductions machines in applications such as wind generation, electric vehicles, or automated processes in which the machine works most of time under transient conditions. For these cases, new diagnostic methodologies have been proposed, based on the use of advanced time-frequency transforms—as, for example, the continuous wavelet transform, the Wigner Ville distribution, or the analytic function based on the Hilbert transform—which enables to track the fault components evolution along time. All these transforms have high computational costs and, furthermore, generate as results complex spectrograms, which require to be interpreted for qualified technical staff. This paper introduces a new methodology for the diagnosis of faults of IM working in transient conditions, which, unlike the methods developed up to today, analyzes the current signal in the slip-instantaneous frequency plane (s-IF), instead of the time-frequency (t-f) plane. It is shown that, in the s-IF plane, the fault components follow patterns that that are simple and unique for each type of fault, and thus does not depend on the way in which load and speed vary during the transient functioning; this characteristic makes the diagnostic task easier and more reliable. This work introduces a general scheme for the IMs diagnostic under transient conditions, through the analysis of the stator current in the s-IF plane. Another contribution of this paper is the introduction of the specific s-IF patterns associated with three different types of faults (rotor asymmetry fault, mixed eccentricity fault, and single-point bearing defects) that are theoretically justified and experimentally tested. As the calculation of the IF of the fault component is a key issue of the proposed diagnostic method, this paper also includes a comparative analysis of three different mathematical tools for calculating the IF, which are compared not only theoretically but also experimentally, comparing their performance when are applied to the tested diagnostic signals.

This paper is about designing a 3D no n-stationary wideband indoor channel model for radio-frequency sensing. The proposed channel model allows for simulating the time-variant (TV) characteristics of the received signal of indoor channel in the presence of a moving object. The moving object is modelled by a point scatterer which travels along a trajectory. The trajectory is described by the object’s TV speed, TV horizontal angle of motion, and TV vertical angle of motion. An expression of the TV Doppler frequency caused by the moving scatterer is derived. Furthermore, an expression of the TV complex channel transfer function (CTF) of the received signal is provided, which accounts for the influence of a moving object and fixed objects, such as walls, ceiling, and furniture. An approximate analytical solution of the spectrogram of the CTF is derived. The proposed channel model is confirmed by measurements obtained from a pendulum experiment. In the pendulum experiment, the trajectory of the pendulum has been measured by using an inertial-measurement unit (IMU) and simultaneously collecting CSI data. For validation, we have compared the spectrogram of the proposed channel model fed with IMU data with the spectrogram characteristics of the measured CSI data. The proposed channel model paves the way towards designing simulation-based activity recognition systems.

Interference detection is very important for Global Navigation Satellite System (GNSS) receivers. Current work on interference detection in GNSS receivers has mainly focused on time-frequency (TF) analysis techniques, such as spectrogram and Wigner–Ville distribution (WVD), where the spectrogram approach presents the TF resolution trade-off problem, since the analysis window is used, and the WVD method suffers from the very serious cross-term problem, due to its quadratic TF distribution nature. In order to solve the cross-term problem and to preserve good TF resolution in the TF plane at the same time, in this paper, a new TF distribution by using a reassigned spectrogram has been proposed in interference detection for GNSS receivers. This proposed reassigned spectrogram method efficiently combines the elimination of the cross-term provided by the spectrogram itself according to its inherent nature and the improvement of the TF aggregation property achieved by the reassignment method. Moreover, a notch filter has been adopted in interference mitigation for GNSS receivers, where receiver operating characteristics (ROCs) are used as metrics for the characterization of interference mitigation performance. The proposed interference detection method by using a reassigned spectrogram is evaluated by experiments on GPS L1 signals in the disturbing scenarios in comparison to the state-of-the-art TF analysis approaches. The analysis results show that the proposed interference detection technique effectively overcomes the cross-term problem and also keeps good TF localization properties, which has been proven to be valid and effective to enhance the interference Sensors 2015, 15 22168 detection performance; in addition, the adoption of the notch filter in interference mitigation has shown a significant acquisition performance improvement in terms of ROC curves for GNSS receivers in jamming environments.

The classification of emotional speech is mostly considered in speech-related research on human-computer interaction (HCI). In this paper, the purpose is to present a novel feature extraction based on multi-resolutions texture image information (MRTII). The MRTII feature set is derived from multi-resolution texture analysis for characterization and classification of different emotions in a speech signal. The motivation is that we have to consider emotions have different intensity values in different frequency bands. In terms of human visual perceptual, the texture property on multi-resolution of emotional speech spectrogram should be a good feature set for emotion classification in speech. Furthermore, the multi-resolution analysis on texture can give a clearer discrimination between each emotion than uniform-resolution analysis on texture. In order to provide high accuracy of emotional discrimination especially in real-life, an acoustic activity detection (AAD) algorithm must be applied into the MRTII-based feature extraction. Considering the presence of many blended emotions in real life, in this paper make use of two corpora of naturally-occurring dialogs recorded in real-life call centers. Compared with the traditional Mel-scale Frequency Cepstral Coefficients (MFCC) and the state-of-the-art features, the MRTII features also can improve the correct classification rates of proposed systems among different language databases. Experimental results show that the proposed MRTII-based feature information inspired by human visual perception of the spectrogram image can provide significant classification for real-life emotional recognition in speech.

In this paper, we present a novel texture image feature for Emotion Sensing in Speech (ESS). This idea is based on the fact that the texture images carry emotion-related information. The feature extraction is derived from time-frequency representation of spectrogram images. First, we transform the spectrogram as a recognizable image. Next, we use a cubic curve to enhance the image contrast. Then, the texture image information (TII) derived from the spectrogram image can be extracted by using Laws’ masks to characterize emotional state. In order to evaluate the effectiveness of the proposed emotion recognition in different languages, we use two open emotional databases including the Berlin Emotional Speech Database (EMO-DB) and eNTERFACE corpus and one self-recorded database (KHUSC-EmoDB), to evaluate the performance cross-corpora. The results of the proposed ESS system are presented using support vector machine (SVM) as a classifier. Experimental results show that the proposed TII-based feature extraction inspired by visual perception can provide significant classification for ESS systems. The two-dimensional (2-D) TII feature can provide the discrimination between different emotions in visual expressions except for the conveyance pitch and formant tracks. In addition, the de-noising in 2-D images can be more easily completed than de-noising in 1-D speech.

Speech is the most significant mode of communication among human beings and a potential method for human-computer interaction (HCI) by using a microphone sensor. Quantifiable emotion recognition using these sensors from speech signals is an emerging area of research in HCI, which applies to multiple applications such as human-reboot interaction, virtual reality, behavior assessment, healthcare, and emergency call centers to determine the speaker’s emotional state from an individual’s speech. In this paper, we present major contributions for; (i) increasing the accuracy of speech emotion recognition (SER) compared to state of the art and (ii) reducing the computational complexity of the presented SER model. We propose an artificial intelligence-assisted deep stride convolutional neural network (DSCNN) architecture using the plain nets strategy to learn salient and discriminative features from spectrogram of speech signals that are enhanced in prior steps to perform better. Local hidden patterns are learned in convolutional layers with special strides to down-sample the feature maps rather than pooling layer and global discriminative features are learned in fully connected layers. A SoftMax classifier is used for the classification of emotions in speech. The proposed technique is evaluated on Interactive Emotional Dyadic Motion Capture (IEMOCAP) and Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS) datasets to improve accuracy by 7.85% and 4.5%, respectively, with the model size reduced by 34.5 MB. It proves the effectiveness and significance of the proposed SER technique and reveals its applicability in real-world applications.

Due to the existing relationship between microstructural properties and magnetic ones of the ferromagnetic materials, the application potential of the magnetic Barkhausen noise (BN) method to non-destructive testing is constantly growing. However, the stochastic nature of the Barkhausen effect requires the use of advanced signal processing methods. Recently, the need to apply time-frequency (TF) transformations to the processing of BN signals arose. However, various TF methods have been used in the majority of cases for qualitative signal conditioning and no extensive analysis of TF-based information has been conducted so far. Therefore, in this paper, the wide analysis of BN TF representation was carried out. Considering the properties of TF transformations, the Short-Time Fourier Transform (STFT) was used. A procedure for definition of the envelopes of the TF characteristic was proposed. To verify the quality of extracted features, an analysis was performed on the basis of BN signals acquired during stress loading experiments of steel elements. First, the preliminary experiments were processed for various parameters of the measuring system and calculation procedures. The feature extraction procedure was performed for different modes of TF representations. Finally, the distributions of TF features over the loading stages are presented and their information content was validated using commonly used features derived from time T and frequency F domains.

For the elimination of radio-frequency interference (RFI) in a passive microwave radiometer, the threshold level is generally calculated from the mean value and standard deviation. However, a serious problem that can arise is an error in the retrieved brightness temperature from a higher threshold level owing to the presence of RFI. In this paper, we propose a method to detect and mitigate RFI contamination using the threshold level from statistical criteria based on a spectrogram technique. Mean and skewness spectrograms are created from a brightness temperature spectrogram by shifting the 2-D window to discriminate the form of the symmetric distribution as a natural thermal emission signal. From the remaining bins of the mean spectrogram eliminated by RFI-flagged bins in the skewness spectrogram for data captured at 0.1-s intervals, two distribution sides are identically created from the left side of the distribution by changing the standard position of the distribution. Simultaneously, kurtosis calculations from these bins for each symmetric distribution are repeatedly performed to determine the retrieved brightness temperature corresponding to the closest kurtosis value of three. The performance is evaluated using experimental data, and the maximum error and root-mean-square error (RMSE) in the retrieved brightness temperature are served to be less than approximately 3 K and 1.7 K, respectively, from a window with a size of 100 × 100 time–frequency bins according to the RFI levels and cases.

Pneumonia causes the deaths of over a million people worldwide each year, with most occurring in countries with limited access to expensive but effective diagnostic methods, e.g., chest X-rays. Physical examination, the other major established method of diagnosis, suffers from several drawbacks, most notably low accuracy and high interobserver error. We sought to address this diagnostic gap by developing a proof-of-concept non-invasive device to identify the accumulation of fluid in the lungs (consolidation) characteristic of pneumonia. This device, named Tabla after the percussive instrument of the same name, utilizes the technique of auscultatory percussion; a percussive input sound is sent through the chest and recorded with a digital stethoscope for analysis. Tabla analyzes differences in sound transmission through the chest at audible frequencies as a marker for lung consolidation. This paper presents preliminary data from five pneumonia patients and eight healthy subjects. We demonstrate 92.3% accuracy in distinguishing between healthy subjects and patients with pneumonia after data analysis with a K-nearest neighbors algorithm. This prototype device is low cost and simple to implement and may offer a rapid and inexpensive method for pneumonia diagnosis appropriate for general use and in areas with limited medical infrastructure.

Ground conductivity meters use electromagnetic fields for the mapping of geological variations, like the determination of water amount, depending on ground layers, which is important for the state analysis of embankments. The VLF band is contaminated by numerous natural and artificial electromagnetic interference signals. Prior to the determination of ground conductivity, the meter’s working frequency is not possible, due to the variable frequency of the interferences. Frequency management based on the analysis of the selected band using track-before-detect (TBD) algorithms, which allows dynamical frequency changes of the conductivity of the meter transmitting part, is proposed in the paper. Naive maximum value search, spatio-temporal TBD (ST-TBD), Viterbi TBD and a new algorithm that uses combined ST-TBD and Viterbi TBD are compared. Monte Carlo tests are provided for the numerical analysis of the properties for a single interference signal in the considered band, and a new approach based on combined ST-TBD and Viterbi algorithms shows the best performance. The considered algorithms process spectrogram data for the selected band, so DFT (Discrete Fourier Transform) could be applied for the computation of the spectrogram. Real–time properties, related to the latency, are discussed also, and it is shown that TBD algorithms are feasible for real applications.