DCASE 2024 Challenge Task 2 Additional Training Dataset

<Data files will be made accessible on May 15th, 2024.> Description This dataset is the "additional training dataset" for the DCASE 2024 Challenge Task 2. The data consists of the normal/anomalous operating sounds of nine types of real/toy machines. Each recording is a single-channel audio that includes both a machine's operating sound and environmental noise. The duration of recordings varies from 6 to 10 seconds. The following nine types of real/toy machines are used in this task: 3DPrinter AirCompressor BrushlessMotor HairDryer HoveringDrone RoboticArm Scanner ToothBrush ToyCircuit Overview of the task Anomalous sound detection (ASD) is the task of identifying whether the sound emitted from a target machine is normal or anomalous. Automatic detection of mechanical failure is an essential technology in the fourth industrial revolution, which involves artificial-intelligence-based factory automation. Prompt detection of machine anomalies by observing sounds is useful for monitoring the condition of machines. This task is the follow-up from DCASE 2020 Task 2 to DCASE 2023 Task 2. The task this year is to develop an ASD system that meets the following five requirements. 1. Train a model using only normal sound (unsupervised learning scenario) Because anomalies rarely occur and are highly diverse in real-world factories, it can be difficult to collect exhaustive patterns of anomalous sounds. Therefore, the system must detect unknown types of anomalous sounds that are not provided in the training data. This is the same requirement as in the previous tasks. 2. Detect anomalies regardless of domain shifts (domain generalization task) In real-world cases, the operational states of a machine or the environmental noise can change to cause domain shifts. Domain-generalization techniques can be useful for handling domain shifts that occur frequently or are hard-to-notice. In this task, the system is required to use domain-generalization techniques for handling these domain shifts. This requirement is the same as in DCASE 2022 Task 2 and DCASE 2023 Task 2. 3. Train a model for a completely new machine type For a completely new machine type, hyperparameters of the trained model cannot be tuned. Therefore, the system should have the ability to train models without additional hyperparameter tuning. This requirement is the same as in DCASE 2023 Task 2. 4. Train a model using a limited number of machines from its machine type While sounds from multiple machines of the same machine type can be used to enhance the detection performance, it is often the case that only a limited number of machines are available for a machine type. In such a case, the system should be able to train models using a few machines from a machine type. This requirement is the same as in DCASE 2023 Task 2. 5 . Train a model both with or without attribute information While additional attribute information can help enhance the detection performance, we cannot always obtain such information. Therefore, the system must work well both when attribute information is available and when it is not. The last requirement is newly introduced in DCASE 2024 Task2. Definition We first define key terms in this task: "machine type," "section," "source domain," "target domain," and "attributes.". "Machine type" indicates the type of machine, which in the additional training dataset is one of nine: 3D-printer, air compressor, brushless motor, hair dryer, hovering drone, robotic arm, document scanner (scanner), toothbrush, and Toy circuit. A section is defined as a subset of the dataset for calculating performance metrics. The source domain is the domain under which most of the training data and some of the test data were recorded, and the target domain is a different set of domains under which some of the training data and some of the test data were recorded. There are differences between the source and target domains in terms of operating speed, machine load, viscosity, heating temperature, type of environmental noise, signal-to-noise ratio, etc. Attributes are parameters that define states of machines or types of noise. For several machine types, the attributes are hidden. Dataset This dataset consists of nine machine types. For each machine type, one section is provided, and the section is a complete set of training data. A set of test data corresponding to this training data will be provided in another seperate zenodo page as an "evaluation dataset" for the DCASE 2024 Challenge task 2. For each section, this dataset provides (i) 990 clips of normal sounds in the source domain for training and (ii) ten clips of normal sounds in the target domain for training. The source/target domain of each sample is provided. Additionally, the attributes of each sample in the training and test data are provided in the file names and attribute csv files. File names and attribute csv files File names and attribute csv files provide reference labels for each clip. The given reference labels for each training clip include machine type, section index, normal/anomaly information, and attributes regarding the condition other than normal/anomaly. The machine type 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 and the attributes are given by their respective file names. Note that for machine types that has its attribute information hidden, the attribute information in each file names are only labeled as "noAttributes". Attribute csv files are for easy access to attributes that cause domain shifts. In these files, the file names, name of parameters that cause domain shifts (domain shift parameter, dp), and the value or type of these parameters (domain shift value, dv) are listed. Each row takes the following format: [filename (string)], [d1p (string)], [d1v (int | float | string)], [d2p], [d2v]... For machine types that have their attribute information hidden, all columns except the filename column are left blank for each row. Recording procedure Normal/anomalous operating sounds of machines and its related equipment are recorded. Anomalous sounds were collected by deliberately damaging target machines. For simplifying the task, we use only the first channel of multi-channel recordings; all recordings are regarded as single-channel recordings of a fixed microphone. We mixed a target machine sound with environmental noise, and only noisy recordings are provided as training/test data. The environmental noise samples 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. Directory structure - /eval_data - /raw - /3DPrinter - /train (only normal clips) - /section_00_source_train_normal_0001_<attribute>.wav - ... - /section_00_source_train_normal_0990_<attribute>.wav - /section_00_target_train_normal_0001_<attribute>.wav - ... - /section_00_target_train_normal_0010_<attribute>.wav - attributes_00.csv (attribute csv for section 00) - /AirCompressor (The other machine types have the same directory structure as 3DPrinter.) - /BrushlessMotor - /HairDryer - /HoveringDrone - /RoboticArm - /Scanner - /ToothBrush - /ToyCircuit Baseline system The baseline system is available on the Github repository <https://github.com/nttcslab/dcase2023_task2_baseline_ae>. 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. Condition of use This dataset was created jointly by Hitachi, Ltd., NTT Corporation and STMicroelectronics and is available under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) license. Citation <TBD> Contact If there is any problem, please contact us: Tomoya Nishida, [e-mail] Keisuke Imoto, [e-mail] Noboru Harada, [e-mail] Daisuke Niizumi, [e-mail] Yohei Kawaguchi, [e-mail]数据集文件将于2024年5月15日开放获取。 ## 数据集描述 本数据集为DCASE 2024挑战赛任务2的额外训练数据集。数据集包含9类实体/仿真机器的正常与异常运行音频。每条录音均为单声道音频，同时包含机器运行声与环境噪声，录音时长介于6至10秒之间。 本任务涉及的9类机器分别为：3D打印机（3DPrinter）、空气压缩机（AirCompressor）、无刷电机（BrushlessMotor）、电吹风（HairDryer）、悬停无人机（HoveringDrone）、机械臂（RoboticArm）、扫描仪（Scanner）、牙刷（ToothBrush）及玩具电路（ToyCircuit）。 ## 任务概述 异常声音检测（Anomalous Sound Detection, ASD）任务旨在识别目标机器发出的声音属于正常还是异常。机械故障自动检测是第四次工业革命中基于人工智能的工厂自动化的核心技术之一，通过声音快速检测机器异常，有助于实时监控机器运行状态。本任务是DCASE 2020任务2至DCASE 2023任务2的延续，本年度任务要求开发满足以下五项要求的异常声音检测系统： 1. **仅使用正常声音训练模型（无监督学习场景）** 由于现实工厂中异常样本稀缺且类型多样，难以收集全覆盖的异常声音模式，因此系统需能够检测训练数据中未出现的未知类型异常声音。本要求与往届任务一致。 2. **无需考虑域偏移即可检测异常（域泛化任务）** 现实场景中，机器运行状态或环境噪声的变化可能引发域偏移（Domain Shift），域泛化（Domain Generalization）技术可有效应对频繁出现或难以察觉的域偏移问题。本要求与DCASE 2022任务2及DCASE 2023任务2一致。 3. **为全新机器类型训练模型** 针对全新机器类型，无法对训练模型的超参数进行调优，因此系统需具备无需额外超参数调整即可完成模型训练的能力。本要求与DCASE 2023任务2一致。 4. **使用同类型机器的有限样本训练模型** 尽管可通过同类型多台机器的音频提升检测性能，但实际场景中往往仅能获取同类型机器的少量样本。此时系统需能够基于少量同类型机器样本完成模型训练。本要求与DCASE 2023任务2一致。 5. **支持有无属性信息的模型训练** 额外属性信息可提升检测性能，但并非总能获取，因此系统需在属性信息可用与不可用两种场景下均能保持良好性能。本要求为DCASE 2024任务2新增内容。 ## 术语定义 本节将定义本任务中的核心术语：机器类型、片段、源域、目标域及属性。 - **机器类型**：指机器的类别，本额外训练数据集涵盖以下9类：3D打印机、空气压缩机、无刷电机、电吹风、悬停无人机、机械臂、文档扫描仪（Scanner）、牙刷及玩具电路。 - **片段**：指用于计算性能指标的数据集子集。 - **源域**：指录制大部分训练数据与部分测试数据的场景；**目标域**：指录制另一部分训练数据与测试数据的不同场景集合。源域与目标域在运行速度、机器负载、介质粘度、加热温度、环境噪声类型及信噪比等维度存在差异。 - **属性**：指定义机器状态或噪声类型的参数。部分机器类型的属性信息未公开。 ## 数据集详情 本数据集涵盖9类机器，每类机器对应一个片段，该片段包含完整的训练数据。与该训练数据对应的测试集将通过另一独立的Zenodo页面发布，作为DCASE 2024挑战赛任务2的评估数据集。 针对每个片段，本数据集提供：(i) 990条源域正常声音训练样本，(ii) 10条目标域正常声音训练样本。每条样本均标注其所属源域/目标域。此外，训练与测试数据中每条样本的属性信息将通过文件名与属性CSV文件提供。 ## 文件名与属性CSV文件 文件名与属性CSV文件为每条样本提供参考标签。每条训练样本的参考标签包含机器类型、片段索引、正常/异常标签，以及除正常/异常状态外的其他属性信息。机器类型由目录名称标注，片段索引由样本文件名标注。 除评估数据集外，其余数据集的正常/异常标签与属性信息均通过样本文件名标注。请注意，对于属性信息未公开的机器类型，其样本文件名中的属性信息仅标注为`noAttributes`。 属性CSV文件用于便捷获取引发域偏移的属性信息，文件中列出了样本文件名、引发域偏移的参数名称（域偏移参数，Domain Shift Parameter, DP）及该参数的取值或类型（域偏移值，Domain Shift Value, DV）。每行遵循以下格式：`[文件名（字符串）], [d1p（字符串）], [d1v（整数|浮点数|字符串）], [d2p], [d2v]……`。对于属性信息未公开的机器类型，每行除文件名列外其余列均留空。 ## 录制流程 录制机器及相关设备的正常与异常运行声音，异常声音通过故意损坏目标机器采集得到。为简化任务，本数据集仅采用多通道录音的第一声道，所有录音均视为固定麦克风录制的单声道音频。将目标机器声音与环境噪声混合，仅提供带噪的训练与测试录音。环境噪声样本采集自多个真实工厂场景。我们将在投稿截止日期前发布关于本数据集的论文，详细说明录制流程细节。 ## 目录结构 - /eval_data - /raw - /3DPrinter - /train（仅包含正常样本） - /section_00_source_train_normal_0001_<attribute>.wav - …… - /section_00_source_train_normal_0990_<attribute>.wav - /section_00_target_train_normal_0001_<attribute>.wav - …… - /section_00_target_train_normal_0010_<attribute>.wav - attributes_00.csv（片段00的属性CSV文件） - /AirCompressor（其余机器类型的目录结构与3DPrinter一致） - /BrushlessMotor - /HairDryer - /HoveringDrone - /RoboticArm - /Scanner - /ToothBrush - /ToyCircuit ## 基线系统 基线系统可从GitHub仓库<https://github.com/nttcslab/dcase2023_task2_baseline_ae>获取。基线系统提供了简单的入门级方案，在本任务2的数据集上可取得合理的性能表现，尤其适合希望快速熟悉异常声音检测任务的入门研究者。 ## 使用条款 本数据集由日立（Hitachi, Ltd.）、日本电信电话公社（NTT Corporation）及意法半导体（STMicroelectronics）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（Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International, CC BY-NC-SA 4.0）进行授权。 ## 引用格式 <TBD> ## 联系方式 如有任何问题，请联系以下人员： Tomoya Nishida：[e-mail] Keisuke Imoto：[e-mail] Noboru Harada：[e-mail] Daisuke Niizumi：[e-mail] Yohei Kawaguchi：[e-mail]