The Francis Crick Institute
Browse

Fat-U-Net: non-contracting U-Net for free-space optical neural networks

Download (1.37 MB)
poster
posted on 2024-12-18, 11:43 authored by Riad Ibadulla, Thomas M. Chen, Constantino Carlos Reyes-Aldasoro
<p dir="ltr">Modified U-Net optimised for free-space optical neural networks, achieving up 37x speed-up over GPUs, with minimal performance loss in image segmentation.</p><p dir="ltr">Poster presented as part of the Crick BioImage Analysis Symposium 2024.</p><p dir="ltr"><i>Permission has been given by authors to upload to Crick Figshare. Copyright remains with the original authors.</i></p>

History

Related Materials

Usage metrics

    The Francis Crick Institute

    Licence

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC