Raw Waveform Research Articles

Predicting physical fitness levels from resting ECG data: a machine learning approach in cardiovascular assessment

Abstract Background Exercise Stress Testing (EST), an inexpensive noninvasive modality, is widely used for preliminary cardiovascular evaluation and ongoing monitoring of cardiac patients. The MET (Metabolic Equivalent of Task) score from EST, indicating physical fitness, is associated with overall mortality risk and is extremely helpful for preoperative evaluation. Artificial Intelligence (AI) advances, particularly Deep Learning (DL) for resting ECG analysis, have significantly improved cardiac diagnostic accuracy. Objectives This study explores AI's ability to predict MET scores from resting ECGs, aiming to streamline fitness assessments and detect early changes in fitness levels. Methods The current analysis encompasses 45,518 consecutive EST evaluations from 31,532 individual patients from 2017 to 2023. MET scores were automatically computed using an established formula, while interpretations by certified cardiologists provided a benchmark for the AI model. Poor fitness was defined as being within the lowest quartile of MET scores, adjusted for sex and age. A deep learning framework was developed to process a 10-second snippet of resting phase ECG in raw waveform, predicting the likelihood of diminished fitness and thus acting as a binary classifier. The efficacy of this Poor Fitness AI model (PF-AI model) was assessed through metrics such as the area under the receiver operating characteristic curve (AUROC), along with sensitivity, specificity, and accuracy rates. Results The average age of the study group was 55±10 years, predominantly male (68.8%, n = 31,337). Among them, 1,152 were identified as having poor fitness. Applying the model to the entire cohort through 5-fold cross-validation yielded a sensitivity of 78%, a specificity of 81%, and an AUROC of 0.87 in predicting poor fitness. Conclusions Leveraging merely 10 seconds of resting ECG data, the PF-AI model proves to be a robust tool in predicting poor fitness levels, comparable to the outcomes of comprehensive stress testing.

Enhancing analysis of diadochokinetic speech using deep neural networks

Diadochokinetic speech tasks (DDK) involve the repetitive production of consonant-vowel syllables. These tasks are useful in detecting impairments, differential diagnosis, and monitoring progress in speech-motor impairments. However, manual analysis of those tasks is time-consuming, subjective, and provides only a rough picture of speech. This paper presents several deep neural network models working on the raw waveform for the automatic segmentation of stop consonants and vowels from unannotated and untranscribed speech. A deep encoder serves as a features extractor module, replacing conventional signal processing features. In this context, diverse deep learning architectures, such as convolutional neural networks (CNNs) and large self-supervised models like HuBERT, are applied for the extraction process. A decoder model uses derived embeddings to identify frame types. Consequently, the paper studies diverse deep architectures, ranging from linear layers, LSTM, CNN, and transformers. These architectures are assessed for their ability to detect speech rate, sound duration, and boundary locations on a dataset of healthy individuals and an unseen dataset of older individuals with Parkinson’s Disease. The results reveal that an LSTM model performs better than all other models on both datasets and is comparable to trained human annotators.

A general model unifying the adaptive, transient and sustained properties of ON and OFF auditory neural responses.

Sounds are temporal stimuli decomposed into numerous elementary components by the auditory nervous system. For instance, a temporal to spectro-temporal transformation modelling the frequency decomposition performed by the cochlea is a widely adopted first processing step in today's computational models of auditory neural responses. Similarly, increments and decrements in sound intensity (i.e., of the raw waveform itself or of its spectral bands) constitute critical features of the neural code, with high behavioural significance. However, despite the growing attention of the scientific community on auditory OFF responses, their relationship with transient ON, sustained responses and adaptation remains unclear. In this context, we propose a new general model, based on a pair of linear filters, named AdapTrans, that captures both sustained and transient ON and OFF responses into a unifying and easy to expand framework. We demonstrate that filtering audio cochleagrams with AdapTrans permits to accurately render known properties of neural responses measured in different mammal species such as the dependence of OFF responses on the stimulus fall time and on the preceding sound duration. Furthermore, by integrating our framework into gold standard and state-of-the-art machine learning models that predict neural responses from audio stimuli, following a supervised training on a large compilation of electrophysiology datasets (ready-to-deploy PyTorch models and pre-processed datasets shared publicly), we show that AdapTrans systematically improves the prediction accuracy of estimated responses within different cortical areas of the rat and ferret auditory brain. Together, these results motivate the use of our framework for computational and systems neuroscientists willing to increase the plausibility and performances of their models of audition.

A retrospective study of the effects of a vasopressor bolus on systolic slope (dP/dt) and dynamic arterial elastance (Eadyn)

BackgroundTo enhance the utility of functional hemodynamic monitoring, the variables systolic slope (dP/dt) and dynamic arterial elastance (Eadyn) are calculated by the Hypotension Prediction Index (HPI) Acumen® Software. This study was designed to characterize the effects of phenylephrine and ephedrine on dP/dt and Eadyn.MethodsThis was a retrospective, non-randomized analysis of data collected during two clinical studies. All patients required intra-operative controlled mechanical ventilation and had an indwelling radial artery catheter connected to an Acumen IQ sensor. Raw arterial pressure waveform data was downloaded from the patient monitor and all hemodynamic measurements were calculated off-line. The anesthetic record was reviewed for bolus administrations of either phenylephrine or ephedrine. Cardiovascular variables prior to drug administration were compared to those following vasopressor administrations. The primary outcome was the difference for dP/dt and Eadyn at baseline compared with the average after the bolus administration. All data sets demonstrated non-normal distributions so statistical analysis of paired and unpaired data followed the Wilcoxon matched pairs signed-rank test or Mann-Whitney U test, respectively.Results201 doses of phenylephrine and 100 doses of ephedrine were analyzed. All data sets are reported as median [95% CI]. Mean arterial pressure (MAP) increased from 62 [54,68] to 78 [76,80] mmHg following phenylephrine and from 59 [55,62] to 80 [77,83] mmHg following ephedrine. Stroke volume and cardiac output both increased. Stroke volume variation and pulse pressure variation decreased. Both drugs produced significant increases in dP/dt, from 571 [531, 645] to 767 [733, 811] mmHg/sec for phenylephrine and from 537 [509, 596] to 848 [779, 930] mmHg/sec for ephedrine. No significant changes in Eadyn were observed.ConclusionBolus administration of phenylephrine or ephedrine increases dP/dt but does not change Eadyn. dP/dt demonstrates potential for predicting the inotropic response to phenylephrine or ephedrine, providing guidance for the most efficacious vasopressor when treating hypotension.Trial registrationData was collected from two protocols. The first was deemed to not require written, informed consent by the Institutional Review Board (IRB). The second was IRB-approved (Effect of Diastolic Dysfunction on Dynamic Cardiac Monitors) and registered on ClinicalTrials.gov (NCT04177225).

Deriving Automated Device Metadata From Intracranial Pressure Waveforms: A Transforming Research and Clinical Knowledge in Traumatic Brain Injury ICU Physiology Cohort Analysis.

Treatment for intracranial pressure (ICP) has been increasingly informed by machine learning (ML)-derived ICP waveform characteristics. There are gaps, however, in understanding how ICP monitor type may bias waveform characteristics used for these predictive tools since differences between external ventricular drain (EVD) and intraparenchymal monitor (IPM)-derived waveforms have not been well accounted for. We sought to develop a proof-of-concept ML model differentiating ICP waveforms originating from an EVD or IPM. We examined raw ICP waveform data from the ICU physiology cohort within the prospective Transforming Research and Clinical Knowledge in Traumatic Brain Injury multicenter study. Nested patient-wise five-fold cross-validation and group analysis with bagged decision trees (BDT) and linear discriminant analysis were used for feature selection and fair evaluation. Nine patients were kept as unseen hold-outs for further evaluation. ICP waveform data totaling 14,110 hours were included from 82 patients (EVD, 47; IPM, 26; both, 9). Mean age, Glasgow Coma Scale (GCS) total, and GCS motor score upon admission, as well as the presence and amount of midline shift, were similar between groups. The model mean area under the receiver operating characteristic curve (AU-ROC) exceeded 0.874 across all folds. In additional rigorous cluster-based subgroup analysis, targeted at testing the resilience of models to cross-validation with smaller subsets constructed to develop models in one confounder set and test them in another subset, AU-ROC exceeded 0.811. In a similar analysis using propensity score-based rather than cluster-based subgroup analysis, the mean AU-ROC exceeded 0.827. Of 842 extracted ICP features, 62 were invariant within every analysis, representing the most accurate and robust differences between ICP monitor types. For the nine patient hold-outs, an AU-ROC of 0.826 was obtained using BDT. The developed proof-of-concept ML model identified differences in EVD- and IPM-derived ICP signals, which can provide missing contextual data for large-scale retrospective datasets, prevent bias in computational models that ingest ICP data indiscriminately, and control for confounding using our model's output as a propensity score by to adjust for the monitoring method that was clinically indicated. Furthermore, the invariant features may be leveraged as ICP features for anomaly detection.

Deep learning approach for automatic speech recognition in the presence of noise

Accurate recognition of speech in noisy environment is still an obstacle for wider application of speech recognition technology. The robustness of a speech recognition system is heavily influenced by the ability to handle the presence of background noise. In this research work, we propose a model based on Deep Fourier Neural Network (DFNN) for Automatic Speech Recognition (ASR) using LibriSpeech dataset. Most of the existing speech recognition techniques lack the robustness of handling background noise, as a result these techniques are not applicable in real-time. In order to mitigate the challenges of background noise, this research work proposes an efficient recognition technique which analyses in detail the raw audio waveforms using the Deep Fourier Neural Network (DFNN). This novel deep learning approach has a concise architecture and is an efficient approach for automatic speech recognition. The proposed deep learning approach embeds the Fourier transform, which is one of the most popular feature representations transform for audio signal processing. The Fourier transform extracts the core information from waveforms in the form of short term spectra of the speech signal as a function of time. The extracted short term spectra are analyzed deeply in the proposed DFNN model for accurate speech recognition in the presence of noise.

Automatic dysarthria detection and severity level assessment using CWT-layered CNN model

Dysarthria is a speech disorder that affects the ability to communicate due to articulation difficulties. This research proposes a novel method for automatic dysarthria detection (ADD) and automatic dysarthria severity level assessment (ADSLA) by using a variable continuous wavelet transform (CWT) layered convolutional neural network (CNN) model. To determine their efficiency, the proposed model is assessed using two distinct corpora, TORGO and UA-Speech, comprising both dysarthria patients and healthy subject speech signals. The research study explores the effectiveness of CWT-layered CNN models that employ different wavelets such as Amor, Morse, and Bump. The study aims to analyze the models’ performance without the need for feature extraction, which could provide deeper insights into the effectiveness of the models in processing complex data. Also, raw waveform modeling preserves the original signal’s integrity and nuance, making it ideal for applications like speech recognition, signal processing, and image processing. Extensive analysis and experimentation have revealed that the Amor wavelet surpasses the Morse and Bump wavelets in accurately representing signal characteristics. The Amor wavelet outperforms the others in terms of signal reconstruction fidelity, noise suppression capabilities, and feature extraction accuracy. The proposed CWT-layered CNN model emphasizes the importance of selecting the appropriate wavelet for signal-processing tasks. The Amor wavelet is a reliable and precise choice for applications. The UA-Speech dataset is crucial for more accurate dysarthria classification. Advanced deep learning techniques can simplify early intervention measures and expedite the diagnosis process.

Fatigue damage identification based on Kullback-Leibler relative entropy for raw acoustic emission waveform

Acoustic emission (AE) as a non-destructive testing technique is used to identify structural fatigue damage through the evolution of AE characteristic parameters, aiming to prevent significant economic losses and catastrophic consequences caused by fatigue failure. Traditional AE characteristic parameters (counts, hit rate, duration, rise time) are susceptible to the influence of threshold and user-defined acquisition parameters (PDT, HDT, HLT). Inaccurate parameter settings can miss the AE signals of the structural damage, resulting in inaccurate assessment. To reduce the dependency on threshold and other user-defined parameters, the Kullback-Leibler relative entropy (KLRE) is proposed as a feature parameter for identifying structural fatigue damage and performing fatigue life analysis. The effectiveness and feasibility of KLRE were verified through the fatigue tension–compression test on medium carbon steel (MCS) 1045 V-notch specimens. Based on different waveform recording methods, KLRE is divided into the hit KLRE (HKLRE) and waveform stream KLRE (SKLRE). Experimental results demonstrate that:(1) SKLRE outperforms the traditional AE time–frequency domain features in assessing fatigue damage, exhibiting an inflection point in its evolution trend before reaching fatigue critical failure. (2) Node-independent wavelet packet denoising method improves the initial instability of SKLRE caused by work hardening and crack face friction. (3) Amplitude attenuation does not affect the evolution trends of Shannon entropy (SE) and SKLRE. (4) The evolution trend of HKLRE is similar to traditional AE time–frequency domain features, necessitating the use of multiple features together to reduce uncertainty.

Augmenting Aquaculture Efficiency through Involutional Neural Networks and Self-Attention for Oplegnathus Punctatus Feeding Intensity Classification from Log Mel Spectrograms.

Understanding the feeding dynamics of aquatic animals is crucial for aquaculture optimization and ecosystem management. This paper proposes a novel framework for analyzing fish feeding behavior based on a fusion of spectrogram-extracted features and deep learning architecture. Raw audio waveforms are first transformed into Log Mel Spectrograms, and a fusion of features such as the Discrete Wavelet Transform, the Gabor filter, the Local Binary Pattern, and the Laplacian High Pass Filter, followed by a well-adapted deep model, is proposed to capture crucial spectral and spectral information that can help distinguish between the various forms of fish feeding behavior. The Involutional Neural Network (INN)-based deep learning model is used for classification, achieving an accuracy of up to 97% across various temporal segments. The proposed methodology is shown to be effective in accurately classifying the feeding intensities of Oplegnathus punctatus, enabling insights pertinent to aquaculture enhancement and ecosystem management. Future work may include additional feature extraction modalities and multi-modal data integration to further our understanding and contribute towards the sustainable management of marine resources.

ARTIFICIAL INTELLIGENT SYSTEM FOR AUTOMATIC DEPRESSION LEVEL ANALYSIS THROUGH VISUAL AND VOCAL EXPRESSIONS

Depression is a major mental health disease of human which is rapidly affecting lives worldwide. Early detection and intervention are crucial for effective treatment and management. Depression is regularly identified with thoughts of suicide. Significant depression can bring about a determination of social and physical side effects. It could remember changes in rest, craving, vitality level, concentration, day by day conduct or confidence. In recent years, deep-learned applications concentrated on neural networks have shown superior performance at hand-crafted apps in various areas. This system presents an innovative artificial intelligence (AI) system designed for automatic depression level analysis by analyzing visual and vocal expressions. Leveraging advances in computer vision and natural language processing, the proposed system extracts relevant features from facial expressions and speech patterns to assess depression severity levels accurately. Deep-learned apps that settle the above issues that may precisely assess the degree of voice and face depression. In the proposed method, Convolutionary Neural Networks (CNN) is developed for learning deep-learned features and descriptive raw waveforms for visual expressions. Second. Keywords: Artificial Intelligence, Depression, Vocal, Facial Expressions, Deep Learning

WaveVC: Speech and Fundamental Frequency Consistent Raw Audio Voice Conversion

Voice conversion (VC) is a task for changing the speech of a source speaker to the target voice while preserving linguistic information of the source speech. The existing VC methods typically use mel-spectrogram as both input and output, so a separate vocoder is required to transform mel-spectrogram into waveform. Therefore, the VC performance varies depending on the vocoder performance, and noisy speech can be generated due to problems such as train-test mismatch. In this paper, we propose a speech and fundamental frequency consistent raw audio voice conversion method called WaveVC. Unlike other methods, WaveVC does not require a separate vocoder and can perform VC directly on raw audio waveform using 1D convolution. This eliminates the issue of performance degradation caused by the train-test mismatch of the vocoder. In the training phase, WaveVC employs speech loss and F0 loss to preserve the content of the source speech and generate F0 consistent speech using the pre-trained networks. WaveVC is capable of converting voices while maintaining consistency in speech and fundamental frequency. In the test phase, the F0 feature of the source speech is concatenated with a content embedding vector to ensure the converted speech follows the fundamental frequency flow of the source speech. WaveVC achieves higher performances than baseline methods in both many-to-many VC and any-to-any VC. The converted samples are available online.

Automatic Identification of Emergency Sounds for Self-Driving Cars

As the development of Self-Driving Cars (SDCs) advances, one important feature that requires attention is their ability to sense and respond to emergencies on the road. Drivers of emergency vehicles prioritize speed over safety during emergencies to save lives, potentially leading to accidents. However, manual cars sometimes may not understand the urgency of the situation, further increasing the risk of collisions. In such cases, self-driving cars offer a better solution to replace manual cars and minimize road accidents. These cars with emergency sound detections offer a level of responsiveness, accuracy, and consistency that surpasses manual cars, contributing to improved road safety and efficiency in emergencies. Therefore, the key objective of the research is to improve the listening capability of self-driving cars to detect emergency vehicles with the help of a hybrid feature extraction technique. The suggested technique leverages a combination of Complex Morlet Wavelet and Co-occurrence Matrix to obtain statistical features from the emergency sounds. The proposed technique can work with the input length of 1.2 seconds of raw waveforms. This research work investigates that self-driving cars can accurately identify emergency vehicles by examining the distinctive emergency sound patterns emitted by the emergency vehicle with the highest accuracy of 94%. At the same time, the proposed technique reduces the computational cost by 20 – 40 milliseconds when compared with other techniques. The result of this work not only provides better accuracy but also reduces detection time, which is a crucial requirement for real-time applications such as self-driving cars.

Crossmixed convolutional neural network for digital speech recognition.

Digital speech recognition is a challenging problem that requires the ability to learn complex signal characteristics such as frequency, pitch, intensity, timbre, and melody, which traditional methods often face issues in recognizing. This article introduces three solutions based on convolutional neural networks (CNN) to solve the problem: 1D-CNN is designed to learn directly from digital data; 2DS-CNN and 2DM-CNN have a more complex architecture, transferring raw waveform into transformed images using Fourier transform to learn essential features. Experimental results on four large data sets, containing 30,000 samples for each, show that the three proposed models achieve superior performance compared to well-known models such as GoogLeNet and AlexNet, with the best accuracy of 95.87%, 99.65%, and 99.76%, respectively. With 5-10% higher performance than other models, the proposed solution has demonstrated the ability to effectively learn features, improve recognition accuracy and speed, and open up the potential for broad applications in virtual assistants, medical recording, and voice commands.

Depth of anesthesia monitoring in Norway-A web-based survey.

The bispectral index (BIS) monitor is the most frequently used electroencephalogram (EEG)-based depth of anesthesia (DoA) technology in Norwegian hospitals. However, there is limited knowledge regarding the extent and clinical impact of its use and how anesthesiologists and nurse anesthetists use the information provided by the DoA monitors in their clinical practice. This cross-sectional survey on the use of DoA monitors in Norway used a web-based questionnaire distributed to anesthesia personnel in all hospitals in Norway. Participation was voluntary and anonymized, and the web form could not track IP sources or respondents' locations. Three hundred and ninety-one nurse anesthetists (n = 324) and anesthesiologists (n = 67) responded. Among the EEG-based DoA monitoring tools, BIS was most often used to observe and assess patients' DoA (98%). Raw EEG waveform analysis (10%), EEG-spectrogram (9%), and suppression rate (10%) were seldom used. Twenty-seven percent of the anesthesia personnel were able to recognize a burst suppression pattern on EEG and its significance. Fifty-eight percent of the respondents considered clinical observations more reliable than BIS. Almost all respondents reported adjusting anesthetic dosage based on the BIS index values (80%). However, the anesthetic dose was more often increased (90%) because of high BIS index values than lowered (55%) because of low BIS index values. Despite our respondents' extensive use of DoA monitoring, the anesthesia personnel in our survey did not use all the information and the potential to guide the titration of anesthetics the DoA monitors provide. Thus, anesthesia personnel could generally benefit from increased knowledge of how EEG-based DoA monitoring can be used to assess and determine individual patients' need for anesthetic medication.

On the role of audio frontends in bird species recognition

Automatic acoustic monitoring of bird populations and their diversity is in demand for conservation planning. This requirement and recent advances in deep learning have inspired sophisticated species recognizers. However, there are still open challenges in creating reliable monitoring systems of natural habitats. One of many open questions is whether predominantly used audio features like mel-filterbanks are appropriate for such analysis since their design follows human's perception of the sound, making them susceptible to discarding fine details from other animals' vocalization. Although research shows that different audio features work better for particular tasks and datasets, it is hard to attribute all advantages to input features since the experimental setups vary. A general solution is to design a learnable audio frontend to extract task-relevant features from raw waveform since it contains all the information in other audio features. The current paper thoroughly analyzes the role of such frontends in bird species recognition, which helped to evaluate the adequacy of traditional time-frequency representations (static frontends) in capturing the relevant information from bird vocalization. In particular, this work shows that the main performance gain in learnable audio frontends comes from the normalization and compression operations rather than the data-driven frequency selectivity and functional form of filters. We observed no significant discrepancy between the frequency bands of the learned and static frontends for bird vocalization. Although the performance of learnable frontends was much higher, we will show that adequate normalization and compression enhance the accuracy of traditional frontends by more than 16% to achieve comparable results for bird species recognition. Ablation studies of the frontends under different configurations and detailed analysis of noise robustness provide evidence for the conclusions, validate the use of mel-filterbanks and similar features in prior works, and provide guidelines for designing future species recognizers. The code is available at https://github.com/houtan-ghaffari/bird-frontends.

SoundCount: Sound Counting from Raw Audio with Dyadic Decomposition Neural Network

In this paper, we study an underexplored, yet important and challenging problem: counting the number of distinct sounds in raw audio characterized by a high degree of polyphonicity. We do so by systematically proposing a novel end-to-end trainable neural network~(which we call DyDecNet, consisting of a dyadic decomposition front-end and backbone network), and quantifying the difficulty level of counting depending on sound polyphonicity. The dyadic decomposition front-end progressively decomposes the raw waveform dyadically along the frequency axis to obtain time-frequency representation in multi-stage, coarse-to-fine manner. Each intermediate waveform convolved by a parent filter is further processed by a pair of child filters that evenly split the parent filter's carried frequency response, with the higher-half child filter encoding the detail and lower-half child filter encoding the approximation. We further introduce an energy gain normalization to normalize sound loudness variance and spectrum overlap, and apply it to each intermediate parent waveform before feeding it to the two child filters. To better quantify sound counting difficulty level, we further design three polyphony-aware metrics: polyphony ratio, max polyphony and mean polyphony. We test DyDecNet on various datasets to show its superiority, and we further show dyadic decomposition network can be used as a general front-end to tackle other acoustic tasks.

Unified Convolutional Sparse Transformer for Disease Diagnosis, Monitoring, Drug Development, and Therapeutic Effect Prediction from EEG Raw Data.

Electroencephalogram (EEG) analysis plays an indispensable role across contemporary medical applications, which encompasses diagnosis, monitoring, drug discovery, and therapeutic assessment. This work puts forth an end-to-end deep learning framework that is uniquely tailored for versatile EEG analysis tasks by directly operating on raw waveform inputs. It aims to address the challenges of manual feature engineering and the neglect of spatial interrelationships in existing methodologies. Specifically, a spatial channel attention module is introduced to emphasize the critical inter-channel dependencies in EEG signals through channel statistics aggregation and multi-layer perceptron operations. Furthermore, a sparse transformer encoder is used to leverage selective sparse attention in order to efficiently process long EEG sequences while reducing computational complexity. Distilling convolutional layers further concatenates the temporal features and retains only the salient patterns. As it was rigorously evaluated on key EEG datasets, our model consistently accomplished a superior performance over the current approaches in detection and classification assignments. By accounting for both spatial and temporal relationships in an end-to-end paradigm, this work facilitates a versatile, automated EEG understanding across diseases, subjects, and objectives through a singular yet customizable architecture. Extensive empirical validation and further architectural refinement may promote broader clinical adoption prospects.

A Lightweight Channel and Time Attention Enhanced 1D CNN Model for Environmental Sound Classification

One dimension convolutional neural networks (1D CNN) that directly take raw waveforms as input has less competition than 2D CNN recognizing environmental sound. In order to overcome its disadvantages, we propose a novel lightweight 1D CNN structure by employing attention mechanism, which has significant improvement in both accuracy and computational complexity. Concretely, (1) two attention modules are constructed along channel and time dimension separately, and combined to give an intermediate feature map, which focus on key frequency band and semantically related time frame information. (2) Without increasing training overhead, snapshot ensemble is employed to further improve performance. Results from two benchmarking datasets (UrbanSound8k, ESC-10) demonstrated that: by employing attention mechanism, our model outperforms all of the previously reported 1D CNN approaches in accuracy with less parameters. Meanwhile with improved performance gain, the proposed model is superior than most of the existing spectral-based 2D CNN approaches and competitive with SOTA performance, while with orders of magnitude parameters fewer. Overall, it indicates our model is compact and has good potential in practical resource-limited applications, such as sound recognition on embedded platform.

Portable cerebral blood flow monitor to detect large vessel occlusion in patients with suspected stroke

BackgroundEarly detection of large vessel occlusion (LVO) facilitates triage to an appropriate stroke center to reduce treatment times and improve outcomes. Prehospital stroke scales are not sufficiently sensitive, so we...

Speaker recognition with global information modelling of raw waveforms

Speaker recognition with global information modelling of raw waveforms