BEGIN:VCALENDAR
METHOD:PUBLISH
PRODID:www-matsim-tf-fau-de//Events//EN
VERSION:2.0
BEGIN:VEVENT
SUMMARY:AI-guided Recognition of Local Patterns in Nanoparticles from 
 Tomographic Images - Yan Mei
UID:cb1c6092-d86a-4eaf-ad08-5f19b8b3a4ae
DESCRIPTION:Yan Mei WW8\, FAU 4. March 2026\, 17:00 WW8\, Fürth This 
 study presents an atom-centred extension of the Bayesian deep-learning
  framework ARISE for the automated structural analysis of nanoparticle
 s reconstructed by atomic electron tomography (AET). By replacing the 
 original box-based strided sampling scheme with spherical\, atom-centr
 ed local environments encoded via SOAP descriptors\, the workflow enab
 les physically interpretable\, uncertainty-aware classification of 108
  crystal prototypes without retraining the pre-trained model. In addit
 ion to prototype probabilities and Bayesian uncertainty estimates\, la
 tent representations from the neural network are analysed using manifo
 ld learning (UMAP) and density-based clustering (HDBSCAN)\, linking le
 arned structural features to real-space atomic configurations. The app
 roach is benchmarked on a synthetic multi-grain polycrystal and subseq
 uently applied to experimental Pd and Pd@Pt nanoparticles\, revealing 
 defect-related distortions and medium-range structural motifs at chemi
 cally diffuse interfaces. Overall\, the study demonstrates that combin
 ing atom-centred descriptors\, Bayesian uncertainty quantification\, a
 nd unsupervised latent-space analysis provides a flexible and data-eff
 icient framework for the exploratory structural characterization of co
 mplex nanomaterials.
DTSTART:20260304T160000Z
DTEND:20260304T170000Z
LOCATION:WW8\, Room 2.018-2\, Dr.-Mack-Str. 77\, Fürth
DTSTAMP:20260505T044017Z
END:VEVENT
END:VCALENDAR