CVMI-Lab / UHDM

(ECCV2022) This is the official PyTorch implementation of ECCV2022 paper: Towards Efficient and Scale-Robust Ultra-High-Definition Image Demoireing
Apache License 2.0
207 stars 28 forks source link

There seems to be something wrong with the compressed file in the dataset download link? #33

Closed DYY512 closed 2 months ago

DYY512 commented 2 months ago

Only the first file "train_split00.tar.gz" can be decompressed, and when decompressing other files, it will show "the file format is unknown or damaged". I don't know if it is my problem or the file's problem?

image

DYY512 commented 2 months ago

The same problem exists for both the training set and the test set. Only the first compressed file can be decompressed, but the others cannot.

XinYu-Andy commented 2 months ago

Hi, it’s not an issue at all. They are special types of compressed files and you need first group them and then uncompress it. Please refer here: https://stackoverflow.com/questions/38199571/untar-multiple-tar-gz-aa-tar-gz-ab-pattern-files

On Thu, 12 Sep 2024 at 01:16, DYY512 @.***> wrote:

Currently, this problem only occurs in the training set. Although the test set also pops up a "file is damaged" prompt, the image can still be decompressed.

— Reply to this email directly, view it on GitHub https://github.com/CVMI-Lab/UHDM/issues/33#issuecomment-2345569730, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOPNNM6WHFVSNS6G7G7Q3MTZWFEULAVCNFSM6AAAAABOCSE3RWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNBVGU3DSNZTGA . You are receiving this because you are subscribed to this thread.Message ID: @.***>

DYY512 commented 2 months ago

Hi, it’s not an issue at all. They are special types of compressed files and you need first group them and then uncompress it. Please refer here: https://stackoverflow.com/questions/38199571/untar-multiple-tar-gz-aa-tar-gz-ab-pattern-files On Thu, 12 Sep 2024 at 01:16, DYY512 @.> wrote: Currently, this problem only occurs in the training set. Although the test set also pops up a "file is damaged" prompt, the image can still be decompressed. — Reply to this email directly, view it on GitHub <#33 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOPNNM6WHFVSNS6G7G7Q3MTZWFEULAVCNFSM6AAAAABOCSE3RWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNBVGU3DSNZTGA . You are receiving this because you are subscribed to this thread.Message ID: @.>

Oh, thank you very much for telling me, otherwise I really don’t know what to do.