BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:SDSC-AI4Science seminar: "Unsupervised Domain Adaptation for Optoa
 coustic Imaging: Leveraging Unpaired Data for Improved Results"
DTSTART:20230425T160000
DTEND:20230425T170000
DTSTAMP:20260510T042406Z
UID:456380c4fbb1e01c256e9d15e881c1135e5da0c5f82ce2352d5d0ee2
CATEGORIES:Conferences - Seminars
DESCRIPTION:Dr. Firat Ozdemir\, Sr. Data Scientist at SDSC\nAbstract\nWi
 th limited data or annotations\, a common practice in AI is to build upon 
 or exploit models pretrained on large-scale natural image datasets and/or 
 generic natural language datasets. However\, in domain sciences\, such as 
 optoacoustic imaging\, it can be challenging to benefit from such pretrain
 ed models. Optoacoustic imaging involves the reconstruction of tissue imag
 es from acoustic signals generated by a light source. As a novel non-invas
 ive tomographic imaging modality\, optoacoustic imaging has a strong poten
 tial. In this talk\, I will describe:\n(i) optoacoustic imaging in simple 
 terms\,\n(ii) some of the challenges that can be tackled with data science
  techniques\,\n(iii) supervised translation with limited-quality paired da
 tasets\,\n(iv) improvements we achieved with unsupervised image-to-image t
 ranslation using unpaired datasets (e.g.\, CycleGAN).\nThis is a joint wor
 k with the Multiscale Functional and Molecular Imaging Lab at ETH Zurich.\
 n\nOrganizers\nThe SDSC - AI4Science seminar is co-organized monthly by th
 e EPFL AI4Science Initiative and the Swiss Data Science Center and foc
 ussing on projects in which data science\, statistics\, machine learning a
 nd AI are applied to the sciences. Each seminar will feature a presentatio
 n of one applied project\, geared towards an audience with expertise in Da
 ta Science methods\, from the initial formulation of a research question i
 n science associated with sources of data\, to the model\, algorithms and 
 analyses produced. The presentation will be highlighting the choices made\
 , the challenges encountered\, interesting technical questions and possibl
 e further developments. A number of the projects presented will be collab
 orative projects of the Swiss Data Science Center. One of the objectives o
 f the seminar is to foster exchanges between researchers working in method
 s and applied data science research in the sciences\, and to create new op
 portunities of collaborations. Each session will feature a talk followed b
 y a discussion with the audience\, to be continued over fingerfood and dri
 nks.
LOCATION:GA 3 21 https://plan.epfl.ch/?room==GA%203%2021 https://epfl.zoom
 .us/j/67864899458
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR