Benchmarking deep learning for automated peak detection on GIWAXS data

© Constantin Völter et al. 2025.

Bibliographische Detailangaben
Veröffentlicht in:Journal of applied crystallography. - 1998. - 58(2025), Pt 2 vom: 01. Apr., Seite 513-522
1. Verfasser: Völter, Constantin (VerfasserIn)
Weitere Verfasser: Starostin, Vladimir, Lapkin, Dmitry, Munteanu, Valentin, Romodin, Mikhail, Hylinski, Maik, Gerlach, Alexander, Hinderhofer, Alexander, Schreiber, Frank
Format: Online-Aufsatz
Sprache:English
Veröffentlicht: 2025
Zugriff auf das übergeordnete Werk:Journal of applied crystallography
Schlagworte:Journal Article Faster R-CNN GIWAXS convolutional neural networks deep learning grazing-incidence wide-angle X-ray scattering peak detection
LEADER 01000caa a22002652c 4500
001 NLM386453381
003 DE-627
005 20250509101626.0
007 cr uuu---uuuuu
008 250508s2025 xx |||||o 00| ||eng c
024 7 |a 10.1107/S1600576725000974  |2 doi 
028 5 2 |a pubmed25n1363.xml 
035 |a (DE-627)NLM386453381 
035 |a (NLM)40170972 
040 |a DE-627  |b ger  |c DE-627  |e rakwb 
041 |a eng 
100 1 |a Völter, Constantin  |e verfasserin  |4 aut 
245 1 0 |a Benchmarking deep learning for automated peak detection on GIWAXS data 
264 1 |c 2025 
336 |a Text  |b txt  |2 rdacontent 
337 |a ƒaComputermedien  |b c  |2 rdamedia 
338 |a ƒa Online-Ressource  |b cr  |2 rdacarrier 
500 |a Date Revised 03.04.2025 
500 |a published: Electronic-eCollection 
500 |a Citation Status PubMed-not-MEDLINE 
520 |a © Constantin Völter et al. 2025. 
520 |a Recent advancements in X-ray sources and detectors have dramatically increased data generation, leading to a greater demand for automated data processing. This is particularly relevant for real-time grazing-incidence wide-angle X-ray scattering (GIWAXS) experiments which can produce hundreds of thousands of diffraction images in a single day at a synchrotron beamline. Deep learning (DL)-based peak-detection techniques are becoming prominent in this field, but rigorous benchmarking is essential to evaluate their reliability, identify potential problems, explore avenues for improvement and build confidence among researchers for seamless integration into their workflows. However, the systematic evaluation of these techniques has been hampered by the lack of annotated GIWAXS datasets, standardized metrics and baseline models. To address these challenges, we introduce a comprehensive framework comprising an annotated experimental dataset, physics-informed metrics adapted to the GIWAXS geometry and a competitive baseline - a classical, non-DL peak-detection algorithm optimized on our dataset. Furthermore, we apply our framework to benchmark a recent DL solution trained on simulated data and discover its superior performance compared with our baseline. This analysis not only highlights the effectiveness of DL methods for identifying diffraction peaks but also provides insights for further development of these solutions 
650 4 |a Journal Article 
650 4 |a Faster R-CNN 
650 4 |a GIWAXS 
650 4 |a convolutional neural networks 
650 4 |a deep learning 
650 4 |a grazing-incidence wide-angle X-ray scattering 
650 4 |a peak detection 
700 1 |a Starostin, Vladimir  |e verfasserin  |4 aut 
700 1 |a Lapkin, Dmitry  |e verfasserin  |4 aut 
700 1 |a Munteanu, Valentin  |e verfasserin  |4 aut 
700 1 |a Romodin, Mikhail  |e verfasserin  |4 aut 
700 1 |a Hylinski, Maik  |e verfasserin  |4 aut 
700 1 |a Gerlach, Alexander  |e verfasserin  |4 aut 
700 1 |a Hinderhofer, Alexander  |e verfasserin  |4 aut 
700 1 |a Schreiber, Frank  |e verfasserin  |4 aut 
773 0 8 |i Enthalten in  |t Journal of applied crystallography  |d 1998  |g 58(2025), Pt 2 vom: 01. Apr., Seite 513-522  |w (DE-627)NLM098121561  |x 0021-8898  |7 nnas 
773 1 8 |g volume:58  |g year:2025  |g number:Pt 2  |g day:01  |g month:04  |g pages:513-522 
856 4 0 |u http://dx.doi.org/10.1107/S1600576725000974  |3 Volltext 
912 |a GBV_USEFLAG_A 
912 |a SYSFLAG_A 
912 |a GBV_NLM 
912 |a GBV_ILN_350 
951 |a AR 
952 |d 58  |j 2025  |e Pt 2  |b 01  |c 04  |h 513-522