DCASE 2021 Challenge Task 2 Development Dataset

<strong>Description</strong> This dataset is the "development dataset" for the <strong>DCASE 2021 Challenge Task 2 "Unsupervised Anomalous Sound Detection for Machine Condition Monitoring under Domain Shifted Conditions"</strong>. The data consists of the normal/anomalous operating sounds of seven types of real/toy machines. Each recording is a single-channel 10-second audio that includes both a machine's operating sound and environmental noise. The following seven types of real/toy machines are used in this task: Fan Gearbox Pump Slide rail ToyCar ToyTrain Valve <strong>Why focus on domain shift?</strong> The task setup of the 2020 version was the ASD under ideal conditions. The training- and testing-phase datasets were generated under the same recording conditions, and enough normal training clips recorded under the test domain were made available. In contrast, real-world cases are more complicated and often involve different machine operating conditions between the training and testing phases. A frequent example of this is when the motor speed continuously varies in a conveyor transporting products on a production line based on the production volume in response to product demand. Since there is infinite variation in rotation speed, the sound will also change with infinite variation. Due to the seasonal demand for many products, a limited period of recording training data limits the motor speed during that period (e.g., 200-300 rpm for autumn) and variations in the training data. However, in the test phase, the ASD system must continue to monitor the conveyor through all seasons, so it must be able to monitor all possible motor speed conditions, including those that differ from the training data (such as 100-400 rpm). In addition to the conditions of the machine, environmental noise conditions (SNR, sound characteristics, etc.) also fluctuate uncontrollably depending on the seasonal demand. In such a situation, the normal state's distribution will be changed (i.e., domain shift).<br> <strong>Definition</strong> First, we define some important terms in this task: "machine type," "section," "source domain," and "target domain." The machine type means the kind of machine, which can be one of seven in this task: fan, gearbox, pump, slide rail, ToyCar, ToyTrain, and valve. The section is defined as a subset of the dataset for calculating performance metrics and is almost identical to what was called "machine ID" in the 2020 version. In the 2020 version, there was a one-to-one correspondence between machine IDs and products, but in the 2021 version, the same product may appear in different sections. Different products may appear in the same section. The source domain means the condition under which most of the training data was recorded, and the target domain means a different condition under which some of the test data was recorded. The source and target domains differ in terms of operating speed, machine load, viscosity, heating temperature, environmental noise, SNR, etc. <strong>Data</strong> This dataset consists of three sections for each machine type (Section 00, 01, and 02), and each section is a complete set of training and test data. For each section, this dataset provides (i) around 1,000 clips of normal sounds in a source domain for training, (ii) only three clips of normal sounds in a target domain for training, (iii) around 100 clips each of normal and anomalous sounds in the source domain for the test, and (iv) around 100 clips each of normal and anomalous sounds in the target domain for the test.<br> <strong>Recording procedure</strong> Normal/anomalous operating sounds of machines and related equipment were recorded. Anomalous sounds were collected by deliberately damaging machines. To simplify the task, we only used the first channel of the multi-channel recordings; all recordings were regarded as single-channel recordings from a fixed microphone. We mixed a machine sound with environmental noise, and only noisy recordings are provided as training/test data. The environmental noise clips were recorded in several real factory environments. We will publish papers on the dataset to explain the details of the recording procedure by the submission deadline.<br> <strong>Reference labels</strong> The given labels for each training/test clip are machine type, section index, normal/anomaly information, and brief attribute information about conditions other than normal/abnormal. The machine type information is given by the directory name. The section index is given by their respective file names. For the datasets other than the evaluation dataset, the normal/anomaly information is given by their respective file names. For the training data, the attribute information is given by their respective file names.<br> <strong>Directory structure</strong> When you unzip the files downloaded from the GitHub repository and Zenodo, you can see the following directory structure. As described in the Dataset section, the machine type information is given by directory name, and the section index, domain, and the condition information are given by file name, as: /dev_data /fan /train (only normal clips) /section_00_source_train_normal_0000_&lt;attribute&gt;.wav ... /section_00_source_train_normal_0999_&lt;attribute&gt;.wav /section_00_target_train_normal_0000_&lt;attribute&gt;.wav /section_00_target_train_normal_0001_&lt;attribute&gt;.wav /section_00_target_train_normal_0002_&lt;attribute&gt;.wav /section_01_source_train_normal_0000_&lt;attribute&gt;.wav ... /section_02_target_train_normal_0999_&lt;attribute&gt;.wav /source_test /section_00_source_test_normal_0000.wav ... /section_00_source_test_normal_0099.wav /section_00_source_test_anomaly_0000.wav ... /section_00_source_test_anomaly_0099.wav /section_01_source_test_normal_0000.wav ... /section_02_source_test_anomaly_0099.wav /target_test /section_00_target_test_normal_0000.wav ... /section_00_target_test_normal_0099.wav /section_00_target_test_anomaly_0000.wav ... /section_00_target_test_anomaly_0099.wav /section_01_target_test_normal_0000.wav ... /section_02_target_test_anomaly_0099.wav /gearbox (The other machine types have the same directory structure as fan.) /pump /slider /ToyCar /ToyTrain /valve The paths of audio files are: "/dev_data/&lt;machine_type&gt;/train/section_[0-9]+_&lt;domain&gt;_train_normal_[0-9]+_&lt;attribute&gt;.wav" "/dev_data/&lt;machine_type&gt;/source_test/section_[0-9]+_source_test_normal_[0-9]+.wav" "/dev_data/&lt;machine_type&gt;/source_test/section_[0-9]+_source_test_anomaly_[0-9]+.wav" "/dev_data/&lt;machine_type&gt;/target_test/section_[0-9]+_target_test_normal_[0-9]+.wav" "/dev_data/&lt;machine_type&gt;/target_test/section_[0-9]+_target_test_anomaly_[0-9]+.wav" For example, the machine type, section, and domain of "/fan/train/section_01_source_train_normal_0108_strenght_1_big_ambient.wav" are "fan", "section 01", and "source", respectively, and its condition is normal. The machine type, section, and domain of "/gearbox/test/section_00_target_test_anomaly_0024.wav" are "gearbox", "section 00", and "target", respectively, and its condition is anomalous. <strong>Baseline system</strong> Two simple baseline systems are available on the Github repository [URL] and [URL]. The baseline systems provide a simple entry-level approach that gives a reasonable performance in the dataset of Task 2. They are good starting points, especially for entry-level researchers who want to get familiar with the anomalous-sound-detection task. <strong>Conditions of use</strong> This dataset was created jointly by <strong>Hitachi, Ltd. </strong>and <strong>NTT Corporation</strong> and is available under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) license. <strong>Publication</strong> If you use this dataset, please cite all the following three papers: Yohei Kawaguchi, Keisuke Imoto, Yuma Koizumi, Noboru Harada, Daisuke Niizumi, Kota Dohi, Ryo Tanabe, Harsh Purohit, and Takashi Endo, "Description and Discussion on DCASE 2021 Challenge Task 2: Unsupervised Anomalous Sound Detection for Machine Condition Monitoring under Domain Shifted Conditions," in arXiv e-prints: 2106.04492, 2021. [URL] Noboru Harada, Daisuke Niizumi, Daiki Takeuchi, Yasunori Ohishi, Masahiro Yasuda, Shoichiro Saito, "ToyADMOS2: Another Dataset of Miniature-Machine Operating Sounds for Anomalous Sound Detection under Domain Shift Conditions," in arXiv e-prints: 2106.02369, 2021. [URL] Ryo Tanabe, Harsh Purohit, Kota Dohi, Takashi Endo, Yuki Nikaido, Toshiki Nakamura, and Yohei Kawaguchi, "MIMII DUE: Sound Dataset for Malfunctioning Industrial Machine Investigation and Inspection with Domain Shifts due to Changes in Operational and Environmental Conditions," in arXiv e-prints: 2105.02702, 2021. [URL] <br> <strong>Feedback</strong> If there is any problem, please contact us: Yohei Kawaguchi, [e-mail] Daisuke Niizumi, [e-mail] Keisuke Imoto, [e-mail]

<strong>数据集描述</strong> 本数据集为<strong>DCASE 2021挑战赛任务2「域偏移条件下面向机器状态监测的无监督异常声音检测（Anomalous Sound Detection, ASD）」</strong>的开发集（development dataset）。数据集包含7类真实/玩具级机器的正常与异常运行声音，每条录音为单声道10秒音频，同时涵盖机器运行声与环境噪声。本次任务使用的7类机器为：风扇(Fan)、齿轮箱(Gearbox)、泵(Pump)、滑轨(Slide rail)、玩具车(ToyCar)、玩具火车(ToyTrain)、阀门(Valve)。
<strong>为何聚焦域偏移？</strong> 2020版任务的设置为理想条件下的异常声音检测，其训练与测试阶段的数据集采用完全一致的录音条件，且可获取测试域下录制的足量正常训练片段。然而现实工业场景更为复杂，训练与测试阶段的机器运行工况往往存在差异。以生产线传送带为例，其电机转速会根据产品需求与产能持续动态调整。由于转速存在无限多种可能，对应的机器声音也会随之产生无限变化。受多数产品的季节性需求影响，训练数据的录制周期有限，因此训练阶段的电机转速仅覆盖某一固定区间（例如秋季为200-300转/分钟），训练数据的分布存在明显局限。但在测试阶段，异常声音检测系统需要全年不间断监测传送带，因此必须能够适配所有可能的电机转速工况，包括与训练数据分布不同的区间（例如100-400转/分钟）。除机器工况外，环境噪声条件（信噪比(Signal-to-Noise Ratio, SNR)、声音特性等）也会随季节需求发生不可控的波动。在此类场景下，正常状态的声音分布会发生显著改变，即域偏移（domain shift）。
<strong>术语定义</strong> 首先对本次任务中的核心术语进行界定：「机器类型」、「段(section)」、「源域(source domain)」与「目标域(target domain)」。机器类型指机器的品类，本次任务中共包含7类：风扇(Fan)、齿轮箱(Gearbox)、泵(Pump)、滑轨(Slide rail)、玩具车(ToyCar)、玩具火车(ToyTrain)与阀门(Valve)。段被定义为用于计算性能指标的数据集子集，与2020版任务中所称的「机器ID(machine ID)」基本一致。2020版任务中，机器ID与产品一一对应，但在2021版任务中，同一产品可能出现在不同段中，不同产品也可共存于同一段中。源域指多数训练数据的录制工况；目标域指部分测试数据的录制工况。二者在运行转速、机器负载、介质粘度、加热温度、环境噪声、信噪比SNR等核心参数上存在显著差异。
<strong>数据</strong> 本数据集针对每类机器类型包含3个段（段00、01与02），每个段均包含完整的训练与测试数据集。针对每个段，数据集提供以下四类数据：(i) 源域下约1000条正常声音片段，用于模型训练；(ii) 目标域下仅3条正常声音片段，用于模型训练；(iii) 源域下约100条正常声音片段与约100条异常声音片段，用于测试集验证；(iv) 目标域下约100条正常声音片段与约100条异常声音片段，用于测试集验证。
<strong>录制流程</strong> 本数据集录制了机器及相关配套设备的正常与异常运行声音。异常声音通过人为可控地损坏机器采集得到。为简化任务复杂度，本次仅使用多通道录音的第一声道，所有录音均视为来自固定麦克风的单声道录音。将机器运行声音与环境噪声混合后，仅提供带噪录音作为训练与测试数据。环境噪声片段录制自多个真实工业工厂环境。我们将在投稿截止日期前发布专门论文，详细说明本数据集的录制流程细节。
<strong>参考标签</strong> 每条训练/测试片段的给定标签包含：机器类型、段索引、正常/异常标识，以及除正常/异常外的工况简要属性信息。其中，机器类型信息通过上级目录名称给出，段索引通过文件名给出。除官方评估集外的所有数据集，其正常/异常标识通过文件名给出；训练数据的属性信息同样通过文件名给出。
<strong>目录结构</strong> 从GitHub仓库与Zenodo平台下载的文件解压后，将得到如下目录结构。如前文数据详情所述，机器类型信息通过目录名称给出，段索引、域类型与工况信息通过文件名给出，具体格式如下：
/dev_data
/fan
/train // 仅包含正常声音片段
/section_00_source_train_normal_0000_&lt;attribute&gt;.wav
...
/section_00_source_train_normal_0999_&lt;attribute&gt;.wav
/section_00_target_train_normal_0000_&lt;attribute&gt;.wav
/section_00_target_train_normal_0001_&lt;attribute&gt;.wav
/section_00_target_train_normal_0002_&lt;attribute&gt;.wav
/section_01_source_train_normal_0000_&lt;attribute&gt;.wav
...
/section_02_target_train_normal_0999_&lt;attribute&gt;.wav
/source_test
/section_00_source_test_normal_0000.wav
...
/section_00_source_test_normal_0099.wav
/section_00_source_test_anomaly_0000.wav
...
/section_00_source_test_anomaly_0099.wav
/section_01_source_test_normal_0000.wav
...
/section_02_source_test_anomaly_0099.wav
/target_test
/section_00_target_test_normal_0000.wav
...
/section_00_target_test_normal_0099.wav
/section_00_target_test_anomaly_0000.wav
...
/section_00_target_test_anomaly_0099.wav
/section_01_target_test_normal_0000.wav
...
/section_02_target_test_anomaly_0099.wav
/gearbox // 其余机器类型的目录结构与风扇一致
/pump
/slider
/ToyCar
/ToyTrain
/valve
音频文件的标准路径格式如下：
"/dev_data/&lt;machine_type&gt;/train/section_[0-9]+_&lt;domain&gt;_train_normal_[0-9]+_&lt;attribute&gt;.wav"
"/dev_data/&lt;machine_type&gt;/source_test/section_[0-9]+_source_test_normal_[0-9]+.wav"
"/dev_data/&lt;machine_type&gt;/source_test/section_[0-9]+_source_test_anomaly_[0-9]+.wav"
"/dev_data/&lt;machine_type&gt;/target_test/section_[0-9]+_target_test_normal_[0-9]+.wav"
"/dev_data/&lt;machine_type&gt;/target_test/section_[0-9]+_target_test_anomaly_[0-9]+.wav"

示例说明：
1. 文件路径"/fan/train/section_01_source_train_normal_0108_strenght_1_big_ambient.wav"对应的机器类型为「风扇」，段为「段01」，域为「源域」，工况为正常。
2. 文件路径"/gearbox/test/section_00_target_test_anomaly_0024.wav"对应的机器类型为「齿轮箱」，段为「段00」，域为「目标域」，工况为异常。
<strong>基线系统</strong> GitHub仓库[URL]与[URL]中提供了两款简易基线系统。该基线方案为入门级的轻量化实现，可在本任务2的数据集上获得合理的性能表现，尤其适合希望快速熟悉异常声音检测任务的入门研究者作为起步参考。
<strong>使用条款</strong> 本数据集由日立有限公司(Hitachi, Ltd.)与日本电信电话株式会社(NTT Corporation)联合开发，采用知识共享署名-非商业性使用-相同方式共享4.0国际许可协议（CC BY-NC-SA 4.0）进行发布。
<strong>引用要求</strong> 若使用本数据集，请务必引用以下三篇论文：
Yohei Kawaguchi, Keisuke Imoto, Yuma Koizumi, Noboru Harada, Daisuke Niizumi, Kota Dohi, Ryo Tanabe, Harsh Purohit, and Takashi Endo, "Description and Discussion on DCASE 2021 Challenge Task 2: Unsupervised Anomalous Sound Detection for Machine Condition Monitoring under Domain Shifted Conditions", in arXiv e-prints: 2106.04492, 2021. [URL]
Noboru Harada, Daisuke Niizumi, Daiki Takeuchi, Yasunori Ohishi, Masahiro Yasuda, Shoichiro Saito, "ToyADMOS2: Another Dataset of Miniature-Machine Operating Sounds for Anomalous Sound Detection under Domain Shift Conditions", in arXiv e-prints: 2106.02369, 2021. [URL]
Ryo Tanabe, Harsh Purohit, Kota Dohi, Takashi Endo, Yuki Nikaido, Toshiki Nakamura, and Yohei Kawaguchi, "MIMII DUE: Sound Dataset for Malfunctioning Industrial Machine Investigation and Inspection with Domain Shifts due to Changes in Operational and Environmental Conditions", in arXiv e-prints: 2105.02702, 2021. [URL]
<strong>反馈与联系</strong> 若存在任何问题，请联系以下人员：
Yohei Kawaguchi, [e-mail]
Daisuke Niizumi, [e-mail]
Keisuke Imoto, [e-mail]