YOLO-WSR: A wavelet feature and structure guided diffusion model for detection and automated restoration of cultural heritage artifact images

Yi Chen, Xing Wang, Wang Shaohua · 2026

Read YOLO-WSR: A wavelet feature and structure guided diffusion model for detection and automated restoration of cultural heritage artifact images by Yi Chen, Xing Wang, Wang Shaohua (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/plos-pone-0351571.txt.

Description

Yi Chen, Xing Wang, and Wang Shaohua published a 2026 methods paper in PLOS ONE on using artificial intelligence to repair images of damaged cultural heritage objects. Their system, called YOLO-WSR, first detects damaged areas in a photograph, then restores them with a diffusion model guided by the object's structure and fine wavelet details. A quality-check module scores each restored image and feeds the result back so the output can be refined again.

On several public datasets, the combined system beat existing approaches at finding damage and produced restorations that scored better on standard image-quality measures. Removing individual parts of the system in turn showed that each piece contributed to the gains. The authors frame the method as practical support for digital preservation work: spotting damage accurately, restoring images automatically, and keeping visual records of fragile heritage objects in usable shape.

Catalog record

CREATORYi Chen, Xing Wang, Wang Shaohua
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE90386
LICENSECC_BY
RIGHTSCC BY
SOURCEPLOS

Source: https://doi.org/10.1371/journal.pone.0351571

Browse the full library · About the Tower of Babel