BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:STEAM Academy Seminar: Topological Deep Learning for the Next Gene
 ration of AI4Science
DTSTART:20260702T123000Z
DTEND:20260702T140000Z
DTSTAMP:20260705T053500Z
UID:indico-event-1696026@indico.cern.ch
CONTACT:steam.academy@cern.ch\;ep-seminars.colloquia@cern.ch
DESCRIPTION:Speakers: Tolga Birdal (Imperial College London)\n\nAbstract: 
 Deep learning transformed artificial intelligence by exploiting structure.
  Convolutions leveraged the geometry of images\, transformers leveraged th
 e structure of sequences\, and graph neural networks enabled learning on r
 elational data. Yet many scientific systems—from particle interactions a
 nd physical fields to molecular assemblies\, cellular processes\, and comp
 lex engineered systems—cannot be faithfully described by pairwise relati
 onships alone. Their behavior emerges from higher-order interactions\, mul
 tiscale organization\, and topological constraints.\nThis talk argues that
  topology is becoming the next organizing principle of AI4Science. Tolga w
 ill introduce Topological Deep Learning\, a rapidly emerging framework tha
 t extends machine learning beyond graphs toward richer topological domains
  capable of representing interactions among groups\, motifs\, cycles\, sur
 faces\, and higher-dimensional structures. He will explore recent advances
  in higher-order message passing\, sheaf learning\, topological neural net
 works\, neural operators\, and transformer architectures under a common pe
 rspective.\nTolga will then demonstrate how these ideas enable new capabil
 ities across scientific discovery\, including molecular foundation models\
 , topology-aware generative models\, learning on biological and physical s
 ystems\, and operator learning for scientific simulation\, presenting an e
 merging scientific ecosystem from challenges and opportunities to open sou
 rce software.\nBio: Dr. Tolga Birdal is an assistant professor (Lecturer) 
 and a UKRI Future Leaders Fellow in the Department of Computing of Imperia
 l College London. Previously\, he was a senior Postdoctoral Research Fello
 w at Stanford University within the Geometric Computing Group of Prof. Leo
 nidas Guibas. Tolga has defended his masters and Ph.D. theses at the Compu
 ter Vision Group under Chair for Computer Aided Medical Procedures\, Techn
 ical University of Munich led by Prof. Nassir Navab. He was also a Doktora
 nd at Siemens AG under the supervision of Dr. Slobodan Ilic working on “
 Geometric Methods for 3D Reconstruction from Large Point Clouds”. His th
 esis was awarded the prestigious EMVA Young Professional Award. His curren
 t foci of interest involve topological / geometric machine learning and 3D
  computer vision. His more theoretical work is aimed at investigating and 
 interrogating limits in geometric computing and non-Euclidean inference as
  well as principles of deep learning. Tolga has several publications at we
 ll-respected venues such as NeurIPS\, CVPR\, ICCV\, ECCV\, ICLR\, T-PAMI\,
  ICRA\, IROS\, ICASSP and 3DV. He is AC for CVPR\, ICCV\, ECCV and is curr
 ently program-chairing 3DV 2025. Aside from his academic life\, Tolga has 
 co-founded multiple companies including Befunky\, a widely used web-based 
 image editing platform.\n \nThis seminar is part of the CERN STEAM Academ
 y Seminar Series. \nA networking coffee break will follow the seminar.\n\
 n\n\nA public webcast will be available and accessible to external partici
 pants.\n\n\n\n \n\nWith the support of CERN's Next Generation Triggers P
 roject.\n\n\n\nhttps://indico.cern.ch/event/1696026/
LOCATION:500/1-001 - Main Auditorium (CERN)
URL:https://indico.cern.ch/event/1696026/
END:VEVENT
END:VCALENDAR