Benchmarking study demonstrates superior performance of DETR over YOLO detectors in Byzantine iconography, highlighting the utility of attention models for religious art analysis.
A novel dataset of Byzantine iconography, ByzArtX, is presented and benchmarked against several state of the art object detectors. The dataset contains images of religious significance from the Byzantine and Christian Art Museum in Athens and Mount Athos, Greece. The icon images have been manually annotated to contain 124 classes, which represent personas and objects of religious significance. The superior performance of attention-based DETR networks over YOLO-based detectors has been confirmed in our benchmark, consistently with available literature. To the best of the authors knowledge the present dataset is the first application of Computer Vision technology to Eastern Orthodox Christian iconography.
No takes yet. Share an insight, caveat, or question.
Theofilis et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: