shuoli0708
commented
1 week ago the training(test) data is the corrupted single HSI image (think of it as one-shot), but we add group transformation to that single image (defined by the transformation type and group size, e,g. 6). For each image, the neural network is randomly initialized, so Hyper-EI learns to reconstruct in a self-supervised manner.
Sent from Outlook for iOShttps://aka.ms/o0ukef
From: B.J.-La @.> Sent: Thursday, October 3, 2024 10:45:20 PM To: shuoli0708/Hyper-EI @.> Cc: Subscribed @.***> Subject: [shuoli0708/Hyper-EI] How many images been used in the paper, training dataset set size? (Issue #1)
This email was sent to you by someone outside the University. You should only click on links or attachments if you are certain that the email is genuine and the content is safe.
Sorry to bother you, but I wonder how many images have been used in the paper, (i.e training dataset set size)?
— Reply to this email directly, view it on GitHubhttps://github.com/shuoli0708/Hyper-EI/issues/1, or unsubscribehttps://github.com/notifications/unsubscribe-auth/A5DUBCRM43FNJLWNTEXA7RTZZVKABAVCNFSM6AAAAABPJ6RU3GVHI2DSMVQWIX3LMV43ASLTON2WKOZSGU3DIMRVGMZTCNQ. You are receiving this because you are subscribed to this thread.Message ID: @.***>
The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. Is e buidheann carthannais a th’ ann an Oilthigh Dhùn Èideann, clàraichte an Alba, àireamh clàraidh SC005336.
Sorry to bother you, but I wonder how many images have been used in the paper, (i.e training dataset set size)?