Dataset Open Access

ToyADMOS dataset

Yuma Koizumi; Shoichiro Saito; Noboru Harada; Hisashi Uematsu; Keisuke Imoto

Dublin Core Export

<?xml version='1.0' encoding='utf-8'?>
<oai_dc:dc xmlns:dc="" xmlns:oai_dc="" xmlns:xsi="" xsi:schemaLocation="">
  <dc:creator>Yuma Koizumi</dc:creator>
  <dc:creator>Shoichiro Saito</dc:creator>
  <dc:creator>Noboru Harada</dc:creator>
  <dc:creator>Hisashi Uematsu</dc:creator>
  <dc:creator>Keisuke Imoto</dc:creator>
  <dc:description>ToyADMOS dataset is a machine operating sounds dataset of approximately 540 hours of normal machine operating sounds and over 12,000 samples of anomalous sounds collected with four microphones at a 48kHz sampling rate, prepared by Yuma Koizumi and members in NTT Media Intelligence Laboratories. The dataset consists of three sub-dataset: "toy car" for product inspection task, "toy conveyor" for fault diagnosis for fixed machine task, and "toy train" for fault diagnosis for moving machine task.

Since the total size of the ToyADMOS dataset is over 440GB, each sub-dataset is split into 7-9 files by 7-zip (7z-format). The total size of the compressed dataset is approximately 180GB, and that of each sub-dataset is approximately 60GB. Download the zip files corresponding to sub-datasets of interest and use your favorite compression tool to unzip these split zip files.

The detail of the dataset is described in [1] and GitHub: 

License: see the file named LICENSE.pdf

[1] Yuma Koizumi, Shoichiro Saito, Noboru Harada, Hisashi Uematsu and Keisuke Imoto, "ToyADMOS: A Dataset of Miniature-Machine Operating Sounds for Anomalous Sound Detection," in Proc of Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA), 2019. </dc:description>
  <dc:subject>Anomaly detection</dc:subject>
  <dc:subject>Acoustic condition monitoring</dc:subject>
  <dc:title>ToyADMOS dataset</dc:title>
All versions This version
Views 3,3713,377
Downloads 104,938104,924
Data volume 872.0 TB871.9 TB
Unique views 2,9662,971
Unique downloads 12,62812,623


Cite as