VisA (Visual Anomaly)

Other names: VisA

Overview

The dataset contains 12 categories of industrial objects with both normal and anomalous samples. Each category includes RGB images and pixel-level ground truth masks for anomaly segmentation.

Description (from the source)

The dataset contains 12 categories of industrial objects with both normal and anomalous samples. Each category includes RGB images and pixel-level ground truth masks for anomaly segmentation.

Domain
Industrial AI
Data type
Images
Tasks
Anomaly Detection, Anomaly Segmentation
Subfields
Anomaly detection, Quality control
Languages
Not available
Tags
Not available

Statistics and structure

Number of samples
Not available
Number of classes
Not available
Number of features
Not available
Number of subjects / cases
Not available
Number of images
Not available
Number of files
Not available
Size
1.8 GB
Formats
Not available
Missing values
Not available

Size note: Archive size (Content-Length).

Licence and access

Licence
Creative Commons Attribution-NonCommercial-ShareAlike 4.0 (CC-BY-NC-SA-4.0)
Licence documentation
Listed in a catalogue (needs confirmation from the source)
Commercial use
No
Redistribution
Yes
Attribution required
Yes
Access method
Direct download
Access restrictions
Not available
Terms of use
Not available

Non-commercial use only, with attribution and derivatives shared under the same licence. Licence text

Always review the original publisher's terms before use; the explorer grants no licence.

Source and download

Platform
GitHub
Publisher / source
Amazon Science
Identifier at the source
anomalib:visa
DOI
Not available
Metadata source
anomalib
Metadata last updated
2026-11-14

The explorer does not host data files; all links point to the original publisher.

Related papers

  1. Paper that introduced the dataset
    Zou, Y., Jeong, J., Pemula, L., Zhang, D., Dabeer, O. (2022). SPot-the-Difference Self-supervised Pre-training for Anomaly Detection and Segmentation. European Conference on Computer Vision.
    Listed for this dataset in the source catalogue · Evidence

Cite the dataset

Built from recorded fields only; missing fields are omitted, not guessed. It is also good practice to cite the paper that introduced the dataset, if there is one.

APA 7
VisA (Visual Anomaly) [Data set]. (2022). Amazon Science. https://github.com/amazon-science/spot-diff
IEEE
[1] "VisA (Visual Anomaly)," Amazon Science, 2022. [Online]. Available: https://github.com/amazon-science/spot-diff
BibTeX
@misc{visavisualanomaly2022,
  title = {{VisA (Visual Anomaly)}},
  year = {2022},
  publisher = {Amazon Science},
  url = {https://github.com/amazon-science/spot-diff},
  howpublished = {\url{https://github.com/amazon-science/spot-diff}},
  note = {Dataset. Accessed: 2026-11-21}
}

Verification

Verification status
Partially verified
Last link check
Not checked yet
Source verification date
Not available
Verification method
Transcribed from: anomalib 2.7.0 (data/datasets/image/visa.py)
Notes
Fields transcribed from the cited catalog/package file; original URL not yet checked for availability.
Source and check date of each field
FieldSourceDate
descriptionanomalib 2.7.0 (data/datasets/image/visa.py)2026-11-14
original_urlanomalib 2.7.0 (data/datasets/image/visa.py)2026-11-14
licenseanomalib 2.7.0 (data/datasets/image/visa.py)2026-11-14

Similar datasets

Industrial AI Partially verified

BTAD (BeanTech Anomaly Detection)

The dataset contains 3 categories of industrial objects with both normal and anomalous samples. Each category includes RGB images and pixel-level ground truth masks for anomaly…

Samples
Not available
Type
Images
Licence
CC-BY-NC-SA-4.0
Source
University of Udine
Industrial AI Partially verified

MVTec AD

The dataset contains 15 categories of industrial objects with both normal and anomalous samples. Each category includes RGB images and pixel-level ground truth masks for anomaly…

Samples
Not available
Type
Images
Licence
CC-BY-NC-SA-4.0
Source
MVTec
Industrial AI Partially verified

AutoVI (Automotive Visual Inspection)

The AutoVI (Automotive Visual Inspection) dataset is distributed as 6 separate zip files on Zenodo, one per class, with normal training images and test images with…

Samples
Not available
Type
Images
Licence
CC-BY-NC-SA-4.0
Source
Zenodo
Industrial AI Partially verified

MPDD (Metal Parts Defect Detection)

The dataset contains 6 categories of industrial objects (metal parts) with both normal and anomalous samples. Each category includes RGB images and pixel-level ground truth masks…

Samples
Not available
Type
Images
Licence
CC-BY-NC-SA-4.0
Source
Brno University of Technology
Industrial AI Partially verified

MVTec LOCO AD

Dataset for unsupervised anomaly detection and localization with logical constraints (structural and logical anomalies), released by MVTec.

Samples
Not available
Type
Images
Licence
CC-BY-NC-SA-4.0
Source
MVTec
Industrial AI Partially verified

Real-IAD

The dataset contains 30 categories of industrial objects with both normal and anomalous samples, captured from 5 different camera viewpoints.

Samples
Not available
Type
Images
Licence
CC-BY-NC-SA-4.0
Source
Hugging Face