Monitoring the conditions of machines is vital in the manufacturing industry. Early detection of faulty components in machines for stopping and repairing the failed components can minimize the downtime of the machine. This article presents an approach to detect the failure occurring in drill machines based on drill sounds from Valmet AB. The drill dataset includes three classes: anomalous sounds, normal sounds, and irrelevant sounds, which are also labeled as "Broken", "Normal", and "Other", respectively. Detecting drill failure effectively remains a challenge due to the following reasons. The waveform of drill sound is complex and short for detection. Additionally, in realistic soundscapes, there are sounds and noise in the context at the same time. Moreover, the balanced dataset is small to apply state-of-the-art deep learning techniques. To overcome these aforementioned difficulties, we augmented sounds to increase the number of sounds in the dataset. We then proposed a convolutional neural network (CNN) combined with a long short-term memory (LSTM) to extract features from log-Mel spectrograms and learn global high-level feature representation for the classification of three classes. A leaky rectified linear unit (Leaky ReLU) was utilized as the activation function for our proposed CNN instead of the rectified linear unit (ReLU). Moreover, we deployed an attention mechanism at the frame level after the LSTM layer to learn long-term global feature representations. As a result, the proposed method reached an overall accuracy of 92.35% for the drill failure detection system.
DataDRILL: Formation Pressure Prediction and Kick Detection for Drilling Rigs
National Subsea Centre（国家海底中心）；Robert Gordon University（罗伯特戈登大学）；ITMO University（ITMO大学）；University of Glasgow（格拉斯哥大学）；Federal Research Center of the Russian Academy of Sciences（俄罗斯科学院联邦研究中心）；Energy Transition Institute（能源转型研究所）
A Deep Learning Architecture with Spatio-Temporal Focusing for Detecting Respiratory Anomalies
School of Electronic Engineering and Computer Science, Queen Mary University of London（伦敦玛丽女王大学电子工程与计算机科学学院）；School of Computer Science and Electronic Engineering, University of Essex（埃塞克斯大学计算机科学与电子工程学院）；Center for Digital Safety & Security, Austrian Institute of Technology（奥地利技术研究所数字安全与安保中心）；Amazon Alexa（亚马逊Alexa）；Nuffield Department of Clinical Neurosciences, University of Oxford（牛津大学纳菲尔德临床神经科学系）
An Inception-Residual-Based Architecture with Multi-Objective Loss for Detecting Respiratory Anomalies
Queen Mary University of London（伦敦玛丽女王大学）；University of Essex（埃塞克斯大学）；Austrian Institute of Technology（奥地利技术研究所）；Amazon Alexa（亚马逊Alexa）；University of Oxford（牛津大学）
Fruit-CoV: An Efficient Vision-based Framework for Speedy Detection and Diagnosis of SARS-CoV-2 Infections Through Recorded Cough Sounds
Ton Duc Thang University（孙德胜大学）；ASICLAND（ASICLAND公司）；Sejong University（世宗大学）；Hanoi University of Industry（河内工业大学）；Northeastern University（东北大学）；Deakin University（迪肯大学）
Robust sound event detection in bioacoustic sensor networks
Cornell Lab of Ornithology, Cornell University（康奈尔大学康奈尔鸟类学实验室）；Music and Audio Research Laboratory, New York University（纽约大学音乐与音频研究实验室）；Center for Urban Science and Progress, New York University（纽约大学城市科学与进步中心）
Mind the Gap: Detecting Cluster Exits for Robust Local Density-Based Score Normalization in Anomalous Sound Detection
Department of Electronic Systems, Aalborg University（电子系统系，奥尔堡大学）；Pioneer Centre for Artificial Intelligence（先锋人工智能中心）；Mitsubishi Electric Research Laboratories (MERL)（三菱电机研究实验室（MERL））
Detecting COVID-19 from Breathing and Coughing Sounds using Deep Neural Networks
University of Augsburg（奥格斯堡大学）；Imperial College London（帝国理工学院）
COVID-19疫情对全球的影响并不均衡;虽然工业经济体能够生产追踪病毒传播所需的测试，并且大多避免了完全封锁，但发展中国家面临测试能力的问题。在本文中，我们探索了深度学习模型作为一种无处不在的、低成本的预测试方法的使用，用于从移动设备或通过网络获取的呼吸或咳嗽音频记录中检测COVID-19。我们采用了一组卷积神经网络，利用原始的呼吸和咳嗽音频和频谱图来分类说话者是否感染了COVID-19。不同的模型是通过使用贝叶斯优化结合HyperBand的自动超参数调整获得的。所提出的方法优于传统的基线方法的一个很大的利润。最终，它通过集成神经网络实现了74.9%的未加权平均召回率（UAR）或80.7%的ROC曲线下面积（AUC），以严格的受试者独立方式考虑呼吸和咳嗽的最佳测试集结果。单独来看，呼吸声似乎比咳嗽声更适合（76.1% vs 73.7% UAR）。
Inception-Based Network and Multi-Spectrogram Ensemble Applied For Predicting Respiratory Anomalies and Lung Diseases
Austria Institute of Technology（奥地利技术研究所）；Queen Mary University of London（伦敦玛丽女王大学）；University of Lübeck（吕贝克大学）；Singapore Institute of Technology（新加坡理工学院）
A Track-Wise Ensemble Event Independent Network for Polyphonic Sound Event Localization and Detection
University of Chinese Academy of Sciences（中国科学院大学）；University of Surrey（萨里大学）；ByteDance（字节跳动）
复调声音事件定位与检测（SELD）旨在检测具有相应时间活动和空间位置的声音事件类型。本文提出了一种新的数据增强方法--航迹系综事件无关网络。该模型是基于我们以前提出的事件无关网络V2，并扩展了构象块和密集块。针对不同模型间可能出现的航迹排列问题，提出了基于航迹输出格式的航迹集成模型。数据增强方法包含若干数据增强链，其由若干数据增强操作的随机组合组成。该方法还利用对数梅尔频谱图，强度矢量，和空间线索增强对数频谱图（SALSA）的不同模型。我们在L3 DAS 22挑战的任务中评估了我们提出的方法，并获得了排名最高的解决方案，其位置相关的F分数为0.699。源代码已发布。
Sound Event Detection of Weakly Labelled Data with CNN-Transformer and Automatic Threshold Optimization
University of Surrey（萨里大学）；Tencent AI Lab（腾讯人工智能实验室）；Qingdao University of Science and Technology（青岛科技大学）
声音事件检测（SED）是检测音频记录中的声音事件的任务。SED任务的一个挑战是许多数据集，例如声学场景和事件的检测和分类（DCASE）数据集是弱标记的。也就是说，对于每个音频剪辑，只有音频标签，而没有声音事件的开始和偏移时间。\qk{We compare segment-wise and clip-wise training for SED that is lacking in previous works. We propose a convolutional neural network transformer (CNN-Transfomer) for audio tagging and SED, and show that CNN-Transformer performs similarly to a convolutional recurrent neural network (CRNN)}。SED的另一个挑战是检测声音事件需要阈值。以往的工作经验性地设置阈值，并不是一个最佳的方法。为了解决这个问题，我们提出了一种自动阈值优化方法。第一阶段是针对不依赖于阈值的度量（例如平均精度（mAP））来优化系统。第二阶段是针对依赖于这些阈值的度量来优化阈值。我们提出的自动阈值优化系统实现了最先进的音频标记F1为0.646，优于没有阈值优化的0.629，以及声音事件检测F1为0.584，…
Handling Domain Shifts for Anomalous Sound Detection: A Review of DCASE-Related Work
Aalborg University（奥胡斯大学）；Pioneer Centre for AI（先锋人工智能中心）；Nagoya University（名古屋大学）；Kyoto University（京都大学）；Mitsubishi Electric Research Laboratories（三菱电机研究实验室）
Proposal-based Few-shot Sound Event Detection for Speech and Environmental Sounds with Perceivers
Allen Institute for Artificial Intelligence（艾伦人工智能研究所）；Western Washington University（西华盛顿大学）；DocuSign（DocuSign公司）；Pacific Northwest National Laboratory（太平洋西北国家实验室）
Environmental Sound Deepfake Detection Using Deep-Learning Framework
Center for Digital Safety & Security, Austrian Institute of Technology (AIT)（数字安全与安全中心，奥地利技术研究所）；FPT University（FPT大学）；Ton Duc Thang University（-ton Duc Thang大学）；HCM University of Technology（胡志明理工大学）
