8000 GitHub - jeeneo/dejpeg: an image compression removal app (using FBCNN and SCUNet)
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

jeeneo/dejpeg

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

39 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

djpeg_logo

DeJPEG

this is an Android app for removing jpeg compression artifacts and denoising images with FBCNN and SCUNet using (the deprecated) PyTorch Mobile library

features:

  1. batch processing
  2. before/after viewing
  3. exporting as BMP or PNG
  4. different models (FBCNN and SCUNET + grey variants)
  5. no internet (even at the permission level; fully offline, fully private)

download:

apk from releases

models:

FBCNN

SCUNet

screenshot:

Screenshot_20250521_182015_DeJPEG

limitations:

  1. proessed on your phone (no gpu), I'd recommend a device with 4gb of ram or more (the models themselves are ~250mb)
  2. input images above 1000px will be processed in chunks
  3. transparency isnt supported and will be flattened (fix planned)

note:

DeJPEG is not affiliated or related to Topaz's DEJPEG or any other program/project

credits:

@adrianerrea for information on how to create and convert a mobile PyTorch app

FBCNN creators

license:

all models used are under their respective licenses, however this app and its code is under the GPLv3 license, however I ask you not to charge money for this

you are free to embed parts of this app in your own project as long as it remains free/non-paywalled

if you do, credit is appreciated but not required.

0