DCASE 2023 Challenge Task 2 Development Dataset

Description

This dataset is the "development dataset" for the DCASE 2023 Challenge Task 2 "First-Shot Unsupervised Anomalous Sound Detection for Machine Condition Monitoring".

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:

• ToyCar
• ToyTrain
• Fan
• Gearbox
• Bearing
• Slide rail
• Valve

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 2022 Task 2. The task this year is to develop an ASD system that meets the following four 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.

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.

4. Train a model using only one machine from its machine type

While sounds from multiple machines of the same machine type can be used to enhance detection performance, it is often the case that sound data from only one machine are available for a machine type. In such a case, the system should be able to train models using only one machine from a machine type.

The last two requirements are newly introduced in DCASE 2023 Task2 as the "first-shot problem".

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 development dataset is one of seven: fan, gearbox, bearing, slide rail, valve, ToyCar, and ToyTrain.
• 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.

Dataset

This dataset consists of seven machine types. For each machine type, one section is provided, and the section is a complete set of training and test data. For each section, this dataset provides (i) 990 clips of normal sounds in the source domain for training, (ii) ten clips of normal sounds in the target domain for training, and (iii) 100 clips each of normal and anomalous sounds for the test. 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/test 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. 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]...

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

- /dev_data  

    - /raw
        - /fan
            - /train (only normal clips)  
                - /section_00_source_train_normal_0000_<attribute>.wav  
                - ...  
                - /section_00_source_train_normal_0989_<attribute>.wav  
                - /section_00_target_train_normal_0000_<attribute>.wav  
                - ...  
                - /section_00_target_train_normal_0009_<attribute>.wav  
            - /test 
                - /section_00_source_test_normal_0000_<attribute>.wav    
                - ...  
                - /section_00_source_test_normal_0049_<attribute>.wav    
                - /section_00_source_test_anomaly_0000_<attribute>.wav  
                - ...  
                - /section_00_source_test_anomaly_0049_<attribute>.wav  
                - /section_00_target_test_normal_0000_<attribute>.wav
                - ...  
                - /section_00_target_test_normal_0049_<attribute>.wav 
                - /section_00_target_test_anomaly_0000_<attribute>.wav  
                - ...  
                - /section_00_target_test_anomaly_0049_<attribute>.wav 
            - attributes_00.csv (attribute csv for section 00)
    - /gearbox (The other machine types have the same directory structure as fan.)  
    - /bearing
    - /slider (`slider` means "slide rail")
    - /ToyCar  
    - /ToyTrain  
    - /valve  

Baseline system

The baseline system is available on the Github repository 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. and NTT Corporation and is available under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) license.

Citation

If you use this dataset, please cite all the following papers. We will publish a paper on the description of the DCASE 2023 Task 2, so pleasure make sure to cite the paper, too.

• Noboru Harada, Daisuke Niizumi, Yasunori Ohishi, Daiki Takeuchi, and Masahiro Yasuda. First-shot anomaly detection for machine condition monitoring: A domain generalization baseline. In arXiv e-prints: 2303.00455, 2023. [URL]
• Kota Dohi, Tomoya Nishida, Harsh Purohit, Ryo Tanabe, Takashi Endo, Masaaki Yamamoto, Yuki Nikaido, and Yohei Kawaguchi. MIMII DG: sound dataset for malfunctioning industrial machine investigation and inspection for domain generalization task. In Proceedings of the 7th Detection and Classification of Acoustic Scenes and Events 2022 Workshop (DCASE2022), 31-35. Nancy, France, November 2022, . [URL]
• Noboru Harada, Daisuke Niizumi, Daiki Takeuchi, Yasunori Ohishi, Masahiro Yasuda, and Shoichiro Saito. ToyADMOS2: another dataset of miniature-machine operating sounds for anomalous sound detection under domain shift conditions. In Proceedings of the 6th Detection and Classification of Acoustic Scenes and Events 2021 Workshop (DCASE2021), 1–5. Barcelona, Spain, November 2021. [URL]

Contact

If there is any problem, please contact us:

• Kota Dohi, kota.dohi.gr@hitachi.com
• Keisuke Imoto, keisuke.imoto@ieee.org
• Noboru Harada, noboru@ieee.org
• Daisuke Niizumi, daisuke.niizumi.dt@hco.ntt.co.jp
• Yohei Kawaguchi, yohei.kawaguchi.xk@hitachi.com