Detail publikačního výsledku

A deep learning approach for anomaly detection in X-ray images of paintings

MEZINA, A.; BURGET, R.; KOTRLÝ, M.

Originální název

A deep learning approach for anomaly detection in X-ray images of paintings

Anglický název

A deep learning approach for anomaly detection in X-ray images of paintings

Druh

Článek WoS

Originální abstrakt

The intersection of technological advancements and cultural heritage studies has intensified the exploration of historical treasures, captivating historians and enthusiasts alike. Artificial intelligence now plays a key role in forensic art investigations by uncovering hidden patterns to detect forgeries. This study focuses on anomaly detection in X-ray images of paintings using the Ghent Altarpiece for training and testing purposes. We propose a novel model combining a Discriminatively Trained Reconstruction Anomaly Embedding Model (DRAEM), a Nested U-Net, and a new dataset derived from the Altarpiece. The proposed architecture was benchmarked against several state-of-the-art deep learning techniques in anomaly detection. Our model achieved an accuracy of 0.8399 and an F1 score of 0.7869, outperforming other methods in both accuracy and computational efficiency. Results, validated by a domain expert, show strong precision and computational efficiency through semi-supervised learning.

Anglický abstrakt

The intersection of technological advancements and cultural heritage studies has intensified the exploration of historical treasures, captivating historians and enthusiasts alike. Artificial intelligence now plays a key role in forensic art investigations by uncovering hidden patterns to detect forgeries. This study focuses on anomaly detection in X-ray images of paintings using the Ghent Altarpiece for training and testing purposes. We propose a novel model combining a Discriminatively Trained Reconstruction Anomaly Embedding Model (DRAEM), a Nested U-Net, and a new dataset derived from the Altarpiece. The proposed architecture was benchmarked against several state-of-the-art deep learning techniques in anomaly detection. Our model achieved an accuracy of 0.8399 and an F1 score of 0.7869, outperforming other methods in both accuracy and computational efficiency. Results, validated by a domain expert, show strong precision and computational efficiency through semi-supervised learning.

Klíčová slova

Deep Learning, Anomaly detection, X-ray, Paintings, Cultural heritage

Klíčová slova v angličtině

Deep Learning, Anomaly detection, X-ray, Paintings, Cultural heritage

Autoři

MEZINA, A.; BURGET, R.; KOTRLÝ, M.

Rok RIV

2026

Vydáno

02.05.2025

Nakladatel

Springer Nature

Místo

NEW YORK

ISSN

2050-7445

Periodikum

Heritage Science

Svazek

13

Číslo

5

Stát

Spojené království Velké Británie a Severního Irska

Strany od

1

Strany do

11

Strany počet

11

URL

https://www.nature.com/articles/s40494-025-01724-9

Plný text v Digitální knihovně

http://hdl.handle.net/11012/251327

BibTex

@article{BUT197805,
  author="Anzhelika {Mezina} and Radim {Burget} and Marek {Kotrlý}",
  title="A deep learning approach for anomaly detection in X-ray images of paintings",
  journal="Heritage Science",
  year="2025",
  volume="13",
  number="5",
  pages="1--11",
  doi="10.1038/s40494-025-01724-9",
  issn="2050-7445",
  url="https://www.nature.com/articles/s40494-025-01724-9"
}