Interpretation of XRF spectral imaging data using unsupervised machine learning:an investigation into Jusepe de Ribera's Onuphrius the Hermit and its underlying composition
Notice bibliographique
Résumé
Jusepe de Ribera (1591 – 1652) painted the composition Onuphrius the Hermit during the Counter-Reformation period of Naples, when specific religious themes were encouraged and depictions of penitent, suffering saints were regularly produced in Ribera’s workshop [1]. In this painting, Saint Onuphrius is represented with great intensity and extreme realism. The Saint, who chose to live as a hermit in the Egyptian desert, is painted with his hands joined, glancing upwards in prayer. His grey hair and beard, and his emaciated torso reflect his life of struggle and deprivation. Anatomical details are depicted with vibrant brush strokes together with the rendering of intense light, creating an image of reality and extreme spirituality. Multiple versions of this composition exist and a version painted on an oak panel was acquired for the Royal Danish Collection in 1764. When this work was X-radiographed for the first time in 1984, an entirely different composition was discovered below the surface [2]. Turned 90 degrees counter clockwise to a landscape orientation, the underlying composition shows a depiction of the Flight of the Holy Family into Egypt. This first and now hidden composition was likely painted with bright tints of blue, green and red mixed with lead white, while the monochromatic palette of the upper composition mostly contains dense layers of lead white in the figure surrounded by a dark background, which prevent a clear view of the hidden work in the X-radiograph. <br/>A clearer understanding and dating of the underlying composition required determination of which pigments comprised the palettes of the surface versus the hidden motif. Macro-X-ray fluorescence (MA-XRF) spectroscopy was selected as the key non-invasive analytical technique due to its ability to yield chemical information at the elemental level. Moreover, being a scanning method that produces full chemical distribution images, MA-XRF spectroscopy enabled an optimal understanding of the full suite of pigments across the painting in both compositions. MA-XRF elemental mapping was performed at the Conservation and Art Technological Studies (CATS) laboratory of the National Gallery of Denmark (SMK) using the Bruker CRONO system, which has a motorized stage designed to collect chemical distributions across large painted surfaces.<br/>XRF data maps were overlaid onto the X-radiograph and complementary multispectral images to better visualize the hidden painting. A combination of unsupervised machine learning methods based on neural networks [3,4] was used to automatically reduce the XRF spectral imaging data to a set of distinct clusters that share similar spectra, making it possible to identify materials more precisely and deduce the paint sequence. The interpretation was confirmed by additional scanning electron microscopy-energy dispersive X-ray spectroscopy (SEM-EDS) measurements of paint cross sections. Finally, the chemical characterization of the materials together with iconographic research and dendrochronology helped determine the time of execution and the relation between the two compositions in terms of their period of execution.<br/> <br/>[1] E. A. Perez Sanchez, N. Spinoza, Jusepe de Ribera, 1591 – 1652, 1992, 39 – 49. <br/>[2] H. Bjerre, Restoration pictures: an exhibition on the preservation and study of old art, ca. no. 13, 1984, 36 - 37.<br/>[3] Kogou S., Lee L., Shahtahmassebi G., Liang H., X‐Ray Spectrometry, 50(4), 2021, 310-319.<br/>[4] J. T. Machado, A. M. Lopes, Applied Mathematical Modelling, 65, 2019, 614-626.
Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.
Comment cette classification a été obtenuedéplier
Prédiction distillée sur la base complète
Imitation des enseignantsNi prévalence calibrée, ni vérité terrain. Validation humaine à venir. Apprise à partir de 10 348 étiquettes directes de Codex et de 10 348 étiquettes directes de Gemma. Le mode candidate est l'union des têtes enseignantes seuillées; le consensus est leur intersection. Ces sorties portent le statut machine_predicted_unvalidated et ne sont ni des étiquettes humaines ni des étiquettes directes de modèles de pointe.
Scores Codex et Gemma par catégorie
| Catégorie | Codex | Gemma |
|---|---|---|
| Métarecherche | 0,001 | 0,000 |
| Méta-épidémiologie (sens strict) | 0,000 | 0,000 |
| Méta-épidémiologie (sens large) | 0,000 | 0,000 |
| Bibliométrie | 0,000 | 0,000 |
| Études des sciences et des technologies | 0,001 | 0,000 |
| Communication savante | 0,000 | 0,000 |
| Science ouverte | 0,000 | 0,000 |
| Intégrité de la recherche | 0,000 | 0,000 |
| Charge utile insuffisante (le modèle a refusé de juger) | 0,001 | 0,000 |
Scores machine (provisoires)
Les deux têtes enseignantes du modèle étudiant, lues sur ce travail. Un score ordonne la base pour la relecture; il n'affirme jamais une catégorie, et le statut de validation accompagne chaque rangée tel quel.
Scores de référence d'un modèle non mature (critères de maturité non atteints, 7 itérations). Un score ordonne; il n'affirme jamais une catégorie.
score_only:v0-immature-baseline · tel quel depuis la passe de notation : score_only signifie que le nombre peut ordonner les travaux, et qu'aucune étiquette de catégorie n'en découleClassification
machine, non validéePrédiction automatique; un appel candidat d’une seule tête enseignante, pas un consensus.
Le détail, modèle par modèle et score par score, se trouve en fin de page sous « Comment cette classification a été obtenue ».