Info-icon.png We have moved to https://openmodeldb.info/. This new site has a tag and search system, which will make finding the right models for you much easier! If you have any questions, ask here: https://discord.gg/cpAUpDK

Difference between revisions of "Model Database"

From Upscale Wiki
Jump to navigation Jump to search
(Table style test)
 
(283 intermediate revisions by 26 users not shown)
Line 1: Line 1:
{| class="wikitable"
 
|-
 
| '''If a model has no license, that means all rights reserved. You will need a private license or explicit permission from the creator to use it commercially or with modifications. However you may use all model here on non-commercial projects, that may not in any way cost money. If you gain permission to use a model in another way including commercially, you should still credit the author and link to the original model source.'''
 
|}
 
  
=ESRGAN ("old Architecture") Models=
+
{{Note|'''<big><big>We have moved to https://openmodeldb.info/. This new site has a tag and search system, which will make finding the right models for you much easier! If you have any questions, ask here: https://discord.gg/cpAUpDK</big></big>'''}}
  
Models that use the "old" ESRGAN architecture. They can be used either with the official ESRGAN repository (old arch tag) and BasicSR (old arch or preferably victorc's fork). You can also use tools that wrap around one of them, like IEU from Honh.
+
<big><big>If you need to look at the old Model Database anyway, [https://upscale.wiki/w/index.php?title=Model_Database&oldid=1571 look here]</big></big>
 
 
==Image scaling and Video upscaling==
 
 
 
In computer graphics and digital imaging, image scaling refers to the resizing of a digital image. In video technology, the magnification of digital material is known as upscaling or resolution enhancement.
 
 
 
===Drawings===
 
 
 
Drawing is a form of visual art in which a person uses various drawing instruments to mark paper or another two-dimensional medium. Instruments include graphite pencils, pen and ink, various kinds of paints, inked brushes, colored pencils, crayons, charcoal, chalk, pastels, various kinds of erasers, markers, styluses, and various metals (such as silverpoint). Digital drawing is the act of using a computer to draw. Common methods of digital drawing include a stylus or finger on a touchscreen device, stylus- or finger-to-touchpad, or in some cases, a mouse. There are many digital art programs and devices.
 
 
 
====Manga / Anime====
 
 
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://drive.google.com/file/d/1hLhsp3aKurLys4cALQiEkYxyxGWNf1ET/view?usp=drivesdk Deviance V1]
 
|[[User:BlackScout|BlackScout]]
 
|4
 
|[https://www.gnu.org/licenses/gpl-3.0.en.html GNU GPLv3]
 
|Digital Art Upscaler
 
|60K
 
|12
 
|128
 
|5.435K
 
|Digital Art from DeviantArt
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|-
 
|[https://drive.google.com/file/d/1OVqM2D-T5meRBCdFjsGVl0bgQZd8kZe5/view Deviance V2]
 
|[[User:BlackScout|BlackScout]]
 
|4
 
|[https://www.gnu.org/licenses/gpl-3.0.en.html GNU GPLv3]
 
|Digital Art Upscaler
 
|82K
 
|8
 
|128
 
|0.676K
 
|Custom MipMapped
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|-
 
|[https://drive.google.com/open?id=192Mw2_yUwCgqt3tAJ2sRxZ8hYHe4CKrZ FireAlpha]
 
|[[User:BlueAmulet|BlueAmulet]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|Artwork with an alpha channel
 
|1.285M
 
|
 
|128
 
|104k tiles, 1.2k imgs
 
|Fire Emblem artwork with transparency
 
|None
 
|-
 
|[https://1drv.ms/u/s!Aip-EMByJHY20wpGoLuSRzjdqh0T WaifuGAN v3]
 
|[[User:DinJerr|DinJerr]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|Anime / Manga
 
|30K
 
|2
 
|128
 
|0.173K
 
|CG-Painted Anime
 
|Manga109v2
 
|-
 
|[https://drive.google.com/open?id=1Vtwt9kgnmju1edibl07WI6Rg7CrAlHIx Dedither Manga FS]
 
|Jacob_ (buildist)
 
|4
 
|[https://www.gnu.org/licenses/gpl-3.0.en.html GNU GPLv3]
 
|Manga and pixel art
 
|200K
 
|16
 
|128
 
|1,526
 
|Manga109 copied 14 times with different dithering and scaling
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|-
 
|[http://www.mediafire.com/file/w3jujtm752hvdj1/Manga109Attempt.pth.zip/file Manga109Attempt]
 
|Kingdomakrillic
 
|4
 
|[https://creativecommons.org/licenses/by/4.0/ CC BY 4.0]
 
|Anime / Manga
 
|?
 
|4
 
|?
 
|0.1K
 
|Manga109
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|-
 
|[https://drive.google.com/open?id=1uJvdx3g3GEY0VxMnHb0ItBvoc6pmGvuH De-Toon]
 
|[[User:LyonHrt|LyonHrt]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Toon Shading / Sprite
 
|225K
 
|8
 
|128
 
|7.117K
 
|Custom Cartoon-style photos
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|-
 
|[https://drive.google.com/open?id=1bqyG9llxkJ6i6MJTaNUiSgSXtrlPi0d7 Falcon Fanart]
 
|[[User:LyonHrt|LyonHrt]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Anime / Manga
 
|125K
 
|8
 
|128
 
|3.393K
 
|Falcon Fanart
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|-
 
|[https://www23.zippyshare.com/v/lhOStpVa/file.html Fidelbd Pokemodel]
 
|Neo-Raws#4055
 
|2
 
|[http://www.wtfpl.net/ WTFPL]
 
|Old Anime
 
|300K
 
|4
 
|128
 
|10.242K
 
|Pokemon DVD - BluRay
 
|[https://mega.nz/#!vtgSWKQT!K7Asn2zKe4N70R2aV89KEMTKhH3aiyGAAiuQDJF09qs 2xESRGAN]
 
|-
 
|[https://drive.google.com/file/d/114yFJKeYCcr6st7aNNo9FJ8wDbEzLpdz/view?usp=sharing DeviantPixelHD]
 
|Raulsangonzalo
 
|4
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|Digital Art Upscaler
 
|250K
 
|16
 
|128
 
|2.797K
 
|Digital Art from Deviant Art
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|-
 
|[https://1fichier.com/?tsmmdifzgwkwtdfcc5uo Anime Oldies]
 
|solidd93110#7731
 
|4
 
|[http://www.wtfpl.net/ WTFPL]
 
|upscaling old anime
 
|290K
 
|16
 
|128
 
|31.316K
 
|Frames from Anime BluRay
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|-
 
|[https://1fichier.com/?u7kwdxdn6uljha3icte8 Anime Oldies Alternative]
 
|solidd93110#7731
 
|4
 
|[http://www.wtfpl.net/ WTFPL]
 
|captain tsubasa anime
 
|290K
 
|16
 
|128
 
|1.250K
 
|Frames from Anime BluRay
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|-
 
|[https://1fichier.com/?cnw3fwws08tdoaxycimj Sharp Anime V1]
 
|solidd93110#7731
 
|2
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Lines and Details
 
|300K
 
|4
 
|128
 
|7.738K
 
|Anime DVD - BluRay
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 2xPSNR]
 
|-
 
|[https://drive.google.com/open?id=103MX2bvd3GW0MYpC53ABU5VYY6v1t99Y DigiPaint]
 
|TheAtheistGod
 
|4
 
|[https://creativecommons.org/publicdomain/zero/1.0/ CC0]
 
|Digital Art Upscaler
 
|35K
 
|6
 
|128
 
|48.493K
 
|Misc digital art, material studies
 
|4xfalcoon300(manga)
 
|-
 
|[https://u.pcloud.link/publink/show?code=kZb2mFXZOwIwPolF8HYAlU8QE04StpEC9YRX UltraYandere]
 
|[[User:Nmkd|nmkd]]
 
|4
 
|[http://www.wtfpl.net/ WTFPL]
 
|Highly flexible 2D Art upscaling
 
|300k
 
|3
 
|256
 
|4800 unique, 28800 total (augmented)
 
|Yandere4800 (yande.re artworks)
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|}
 
 
 
Manga109Attempt is slightly blurry, but performs well as a general upscaler.
 
 
 
Falcoon Fanart tries to improve upon it with the goal of removing checkerboard patterns / and dithering. It has oil color based shading with sharp lines.
 
 
 
WaifuGAN v3 is [[User:DinJerr|DinJerr]]’s third attempt at training from a mostly anime dataset sourced from image boards and is intended for upscaling CG-painted anime with variable outlines. Only PNGs were used, mainly with brush strokes and gradients. Texturised images avoided as much as possible. If too generative, tone down by interpolating with a softer model.
 
 
 
De-Toon, is a model that does the opposite of tooning an image. It takes toon style shading and detail, and attempts to make it realistic. Its very sensitive, and can be used on small sprites, to large images. Also included is a alt version, which is less sharp.
 
 
 
DigiPaint is a digital art upscaler designed to take brush strokes into account, based off of the Falcoon Fanart model.
 
 
 
DeviantPixelHD was designed to upscale original LucasArts games' backgrounds. High Definition digital art from Deviant Art was used, reducing the LR training to 32 colors so the training images looked pixelated. Can be used with other environments such as manga or digital art. Examples can be found on Raúl Sangonzalo's YouTube account (<nowiki>https://www.youtube.com/channel/UCwBfuiHdSPQ-zslOmiW8OHg</nowiki>).
 
 
 
FireAlpha is an attempt to create a 4 channel model that handles transparency without splitting and upscaling separately.
 
 
 
*'''''Notice:''' IEU can use the model but will attempt to split and handle transparency itself, BlueAmulet's fork of ESRGAN is required to use this model. https://github.com/BlueAmulet/ESRGAN''
 
 
 
4xFSDedither_Manga was intended for pixel art containing dithering patterns, but ended up being effective on pixel art and manga/cartoon style images without any dithering at all.
 
 
 
====Cartoon / Comic====
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://drive.google.com/open?id=1qjpxp8z4FGLZxieUDEVKd6FDum-8vpvJ Comic Book]
 
|[[User:LyonHrt|LyonHrt]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Comic / Drawings
 
|115K
 
|8
 
|128
 
|1.548K
 
|Custom (Spider-Man)
 
|none (no interpolation)
 
|-
 
|[https://mega.nz/file/E1o21QRK#FoXh-5v2KITjc_cjC30HrFoh2etp1cmBkNqruMARPeA DeteriorationFrames 1.0]
 
|Klexos
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Animation
 
|102K
 
|2
 
|128
 
|6.75K
 
|HD to bad DVD transfer
 
| Scratch
 
|-
 
|[https://mega.nz/file/UhZDmYgQ#IzW5VzWV1DKAVZiORhq1ro6RxfmVBBtUFxBPe5X_NRU FilmFrames 0.5]
 
|Klexos
 
|2
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Film Animation
 
|677K
 
|3
 
|192
 
|250
 
|SD to HD film rescan
 
| Scratch
 
|-
 
|[https://mega.nz/#!twxxWQTT!OjPL0o9sq2MM6Flac1shLo9CvE33P9mKKF73OMT-j2I DigitalFrames 2.0]
 
|Klexos
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Digital Animation
 
|905K
 
|27
 
|128
 
|1.636M
 
|Digital Cartoon Images
 
| Scratch
 
|-
 
|[https://mega.nz/#!lwAhQY5D!btyx5fjIRxD6GJwA8Dnnn7beUTvK_NIpDU8Qo69X-Ao DigitalFrames 2.1_Aggressive]
 
|Klexos
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Flat Digital Animation
 
|230K
 
|27
 
|128
 
|1.8M
 
|Digital Cartoon Images
 
|DigitalFrames 2.0
 
|-
 
|[https://drive.google.com/file/d/1RhEkTYy-1DxJP2ynLW5NSrz6BjVMy8xx/view Struzan]
 
|Laserschwert
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|airbrush/pencil-based artwork
 
|300K
 
|4
 
|128
 
|2.790K
 
|movie poster artwork by Drew Struzan and Bill Eaken
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|-
 
|[https://de-next.owncube.com/index.php/s/jYnFtncarkBmpcF FatalimiX]
 
|[[User:Twittman|Twittman]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Comic and Cartoon style images
 
|260K
 
|10
 
|128
 
|69k/79k
 
|Digital Comics
 
|4x_Fatality_MKII_90000_G_02.pth
 
|-
 
|[https://drive.google.com/open?id=1dT7Pw0iwPkQC4-rWriibrVLuA3KS3pS4 SpongeBob]
 
|Joey
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Cartoons & SpongeBob Games
 
|235k
 
|20
 
|128
 
|8k
 
|Scene-detected frames from random Season 11 episodes
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|-
 
|[https://drive.google.com/open?id=1NHuNF36dp_myvPH__IesJ5x6GuppUwS6 SpongeBob v6]
 
|Joey
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Cartoons
 
|190k
 
|20
 
|128
 
|4,803
 
|1 frame of every scene of season 11, downscaled 50% with nearest neighbor
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|-
 
|[https://drive.google.com/open?id=1HFA1e75QvdostGAZVPBxFHtkiru5gLPa Spongebob v6 De-Quantize]
 
|Joey
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Quantized Cartoons
 
|90k
 
|20
 
|128
 
|4,803
 
|1 frame of every scene of season 11, downscaled 50% with nearest neighbor
 
|4x Spongebob v6
 
|-
 
|[https://drive.google.com/open?id=1geNLDAnQzLadMvvoRLhVy4MXhQf5t8JP Spongebob v6 Deblur]
 
|Joey
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Cartoons
 
|65k
 
|20
 
|128
 
|4,803
 
|1 frame of every scene of season 11, downscaled 50% with nearest neighbor
 
|4x Spongebob v6
 
|-
 
|[https://1drv.ms/u/s!Aip-EMByJHY26nuB0Lo0rdirfGoi?e=XI9Fa2 HugePeeps]
 
|[[User:DinJerr|DinJerr]]
 
|8
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|Painted humans
 
|360k
 
|4
 
|256
 
|1,000+
 
|A bunch of pretty girls (and guys) from ArtStation
 
|TGHQFace8x
 
|-
 
|[https://de-next.owncube.com/index.php/s/x99pKzS7TNaErrC Fatal_Anime]
 
|[[User:Twittman|Twittman]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Anime and Cartoons
 
|500k
 
|6
 
|128
 
|183k
 
|Anime and Manga images
 
|4x_Fatality_Comix_MKII_430000_G.pth
 
|-
 
|[https://1drv.ms/u/s!Aip-EMByJHY26wU-wOqApp7Hu7tt?e=4ihnBr HugePaint]
 
|[[User:DinJerr|DinJerr]]
 
|8
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|Digital Illustrations
 
|500k
 
|4
 
|256
 
|1,000+
 
|Variety of images from ArtStation
 
|HugePeeps
 
|}
 
 
 
The Comic Book model was trained using stills from the film spiderman into the spiderverse, has a comic book crosshatch shading effect to the images. [https://cdn.discordapp.com/attachments/547950274313191425/564427623204388864/6de7c88f.png Sample]
 
 
 
FilmFrames 0.5: Purpose of the model is upscaling medium quality sourced SD animation frames to an HD "Film rescan" status. https://imgsli.com/MjA2NDM
 
 
 
DeteriorationFrames 1.0: The purpose of this model is to aid in the creation of datasets aimed at animation by saving you time and allowing you to deteriorate the HQ/HD frames directly though this model instead of having to go through frame by frame to find SD/LQ counterparts. This iteration in particular adds dot crawl, ringing and chromatic aberrations. https://imgsli.com/MTk4MjA.
 
Important note: Resizing your HD counterparts before using this model to create the LRs is essential.
 
Tip: If you blur the HRs after resizing them but before using the deterioration model on them, depending on how much you blur you'll get less Chromatic aberrations. So with this you can tweak both blurriness and the aberrations to match better the source you intend to actually use the model you are about to train on.
 
 
 
DigitalFrames 2.0: Was trained completely separately from 1.0 with a completely new and bigger dataset. It has a reduced sensitivity to noise which reduces the blobs of noise phenomena. It can also work reasonably well on Cel/Film Toons now. Overall it's a big leap from 1.0.
 
 
 
DigitalFrames 2.1_Aggressive: Provides the sharpest results yet. It's the most aggressive version of DF. Does it's best on animation with little to no textures. The model as well as any version of DF is not suitable for all kinds of animation.
 
 
 
FatalimiX is specialized in making those low res comics high res again!
 
 
 
The SpongeBob model was trained for the purpose of upscaling Battle for Bikini Bottom's textures, but it works well for upscaling scenes from the show as well. It was initially trained on the clean downscaled images but later JPEG, quantization, dithering, and Gaussian blur were added to the LR OTF options to help it upscale the textures better.
 
 
 
Spongebob v6 is ideally a lot sharper and cleaner, but Joey's still not sure if it works better than the old one. It still seems to be better in many cases.
 
 
 
Spongebob v6 De-Quantize is for de-quantizing as well as upscaling. The results are pretty blurry but it works decently enough.
 
 
 
Spongebob v6 Deblur was created as Joey wasn't entirely happy with the main Spongebob v6 model. It was trained with blurring OTF options and two different downscale types.
 
 
 
HugePeeps v1 was created by DinJerr to test out a theory on how to properly train 8x models. This particular model best works on painted pics of ladies.
 
 
 
HugePaint v1 is trained on a more generic dataset with a balance of illustrated humans, vegetation, minerals and some animals.
 
 
 
Some of the first cartoon painted models for ESRGAN can still be found [https://esrgan.blogspot.com/2019/01 here], but have been mostly superseded by newer models.
 
 
 
====Pixel Art====
 
 
 
Pixel art is a form of digital art, created through the use of software, where images are edited on the pixel level. The aesthetic for this kind of graphics comes from 8-bit and 16-bit computers and video game consoles, in addition to other limited systems such as graphing calculators. In most pixel art, the color palette used is extremely limited in size, with some pixel art using only two colors.
 
 
 
Creating or modifying pixel art characters or objects for video games is sometimes called spriting, a term that arose from the hobbyist community. The term likely came from the term sprite, a term used in computer graphics to describe a two-dimensional bitmap that is used in tandem with other bitmaps to construct a larger scene.
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://1drv.ms/u/s!Aip-EMByJHY27CHg3-1Uue5KgbtK?e=6k0COU BigFace v3]
 
|[[User:DinJerr|DinJerr]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|Painted people
 
|300K~
 
|4
 
|384
 
|1K
 
|Selected concept art
 
|[https://1drv.ms/u/s!Aip-EMByJHY23ii1LoVMFNVOprBG?e=u2YxRU BigFace]
 
|-
 
|[https://1drv.ms/u/s!Aip-EMByJHY200v-vXPezsCsUd5T?e=r7mBNn BigFArt Suite]
 
|[[User:DinJerr|DinJerr]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|Larger-scaled pixels to digital painting
 
|800+
 
|6
 
|192
 
|~12K
 
|DIV2K, FatalityFaces, ArtStation paintings
 
|Face Focus
 
|-
 
|[https://1drv.ms/u/s!Aip-EMByJHY22BhfdJ1Epkfc3ugq?e=yVfXkz FArtDIV3 Suite]
 
|[[User:DinJerr|DinJerr]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|Pixels to digital painting
 
|700K~
 
|4
 
|128
 
|~10K
 
|DIV2K, FatalityFaces, ArtStation paintings
 
|ArtStation1337
 
|-
 
|[https://drive.google.com/open?id=1BzT-KwUBiZzACxGdIN6w930aOVvvfhg7 Faithful]
 
|Joey
 
|2
 
|[https://creativecommons.org/publicdomain/zero/1.0/ CC0]
 
|General small pixel textures
 
|130K
 
|80
 
|32
 
|2.858K
 
|Faithful 32x32 (HR) / default Minecraft 1.13 and 1.14 (LR)
 
|Private model trained from [https://mega.nz/#!vtgSWKQT!K7Asn2zKe4N70R2aV89KEMTKhH3aiyGAAiuQDJF09qs 2xESRGAN]
 
|-
 
|[https://u.pcloud.link/publink/show?code=XZHawFXZhqD7dF4qtYHnqJVPnQd0ey0vW1JX Faithful Lite]
 
|Joey
 
|2
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|General small pixel textures
 
|275K
 
|128
 
|32
 
|2.928K
 
|Faithful 32x32 1.13-1.16(HR) / default Minecraft 1.13 and 1.16 (LR)
 
|none
 
|-
 
|[https://u.pcloud.link/publink/show?code=kZUMKFXZUG6BpzHR6wFGack19pvWgbQQ9OA7 FakeFaith Lite]
 
|Joey
 
|2
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|General small pixel textures
 
|105K
 
|24
 
|64
 
|24.110K
 
|Spongebob Season 11 frames, downscaled 50% with NN for the HR, and 50% NN again for the LR
 
|[https://u.pcloud.link/publink/show?code=XZHawFXZhqD7dF4qtYHnqJVPnQd0ey0vW1JX Faithful Lite]
 
|-
 
|[https://drive.google.com/open?id=1Uc9RUc2YpZKpPoGQeGxNUb3Ro-U720cH SmoothRealism]
 
|Joey
 
|4
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|Pixel art, rocky/grainy textures
 
|140K
 
|2
 
|64
 
|679
 
|R3DCraft Smooth Realism Texture Pack 64x
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|-
 
|[https://drive.google.com/u/1/uc?id=1WXJQzqVGTKaX_8_Ma1hJ45W5nUkZBGFu&export=download Rebout]
 
|[[User:LyonHrt|LyonHrt]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Character Sprites
 
|325K
 
|8
 
|128
 
|23.808K
 
|Custom prepared sprites from kof 94 rebout, now updated with cx loss
 
|Detoon
 
|-
 
|[https://drive.google.com/u/1/uc?id=1h-bY4gq_61sRF_xK3F4OiVK1cWRX0Q9w&export=download Rebout blend]
 
|[[User:LyonHrt|LyonHrt]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Character Sprites
 
|325K
 
|8
 
|128
 
|23.808K
 
|Custom prepared sprites from kof 94 rebout, which with gradients blending
 
|Detoon
 
|-
 
|[https://drive.google.com/open?id=1jE2le5Lsab-AcMUMm0Zr9L-OvFNhnRAG ScaleNX]
 
|[[User:LyonHrt|LyonHrt]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Scalenx style pixel art upscaler
 
|80K
 
|8
 
|128
 
|1.070K
 
|custom scalenx up-scaled from retroarch shader
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|-
 
|[https://drive.google.com/open?id=1LFd4BqZ2N8p21JzjLX6hGdm7RpVIkH7s Xbrz]
 
|[[User:LyonHrt|LyonHrt]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Xbrz style pixel art upscaler
 
|90K
 
|8
 
|128
 
|1.897K
 
|custom xbrz up-scaled
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|-
 
|[https://drive.google.com/open?id=1VDbjRii9zyyUhyqnQd5H8OP8d3AuwjTe Xbrz+DD]
 
|[[User:LyonHrt|LyonHrt]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Xbrz style pixel art upscaler with de-dithering
 
|90K
 
|8
 
|128
 
|1.523K
 
|custom de-dithered xbrz
 
|xbrz
 
|-
 
|[https://de-next.owncube.com/index.php/s/gPjswdm6gCegQdz Fatality]
 
|[[User:Twittman|Twittman]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|(dithered) spirites
 
|265K
 
|10
 
|128
 
|19.7K
 
|?
 
|[https://de-next.owncube.com/index.php/s/YbAqNMMTtxrajF6 Face]
 
|-
 
|[https://de-next.owncube.com/index.php/s/ECsEHxdoYCnFsZA Fatal Pixels]
 
|[[User:Twittman|Twittman]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Sprites
 
|340K
 
|22
 
|96
 
|49K
 
|Anime, Manga
 
|Fatality_MKII_90k
 
|}
 
 
 
Fatality is meant to be used for upscaling medium resolution Sprites, dithered or un-dithered, it can also upscale manga/anime and gameboy camera images.
 
 
 
Rebout is trained to give detail to character models, with faces and hands improved. Based on the snk game kof94 rebout, although best for snk style games, does work on a variety of sprites. Also included is a interpolated version that may provide a cleaner upscale for certain sprites.
 
 
 
Faithful 2x was trained on the faithful 32x32 Minecraft texture/resource pack, using both the 1.13 and 1.14 versions (which have different textures). For those not familiar, this texture pack is meant to upscale the 16x16 textures to 32x32 faithfully, as in make them as close as possible to the original, but higher resolution. The model does a decent job at replicating this.
 
 
 
*'''NOTE:''' Due to an error in the dataset, sometimes the model turns jagged white lines green. Only encountered this in one texture tried, so it appears unlikely to happen.
 
 
 
FArtDIV3 Suite is a group of models based on a large image dataset trained using OTF tile generation. Most of the models are based off the Base model, and suite different purposes. Fine is used to generate details, but may be noisy/grainy at times. Blend will attempt to blend colour steps into proper gradients. UnholyFArt is a interpolated model that reduces details in favor of sharp blends.
 
 
 
BigFArt Suite is similar to FArtDIV3, but for larger pixel sprites/art in the context of tile size.
 
 
 
BigFace_v3 Suite takes the BigFace model from the BigFArt suite, and tries to improve by using double-sized tiles. Main model has good dedithering. _Blend focuses on sharp blends, and _Clear will abstract and sharpen details.
 
 
 
===Photographs and Photorealism===
 
 
 
A photograph (also known as a photo) is an image created by light falling on a photosensitive surface, usually photographic film or an electronic image sensor, such as a CCD or a CMOS chip. Most photographs are created using a camera, which uses a lens to focus the scene's visible wavelengths of light into a reproduction of what the human eye would see. The process and practice of creating such images is called photography. The word photograph was coined in 1839 by Sir John Herschel and is based on the Greek φῶς (phos), meaning "light," and γραφή (graphê), meaning "drawing, writing," together meaning "drawing with light."
 
Photorealism is a genre of art that encompasses painting, drawing and other graphic media, in which an artist studies a photograph and then attempts to reproduce the image as realistically as possible in another medium. Although the term can be used broadly to describe artworks in many different media, it is also used to refer specifically to a group of paintings and painters of the American art movement that began in the late 1960s and early 1970s.
 
 
 
====Misc / Kitchen Sink====
 
 
 
All kinds of photographs or photorealistic images. Those models aren't specialized.
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://e.pcloud.link/publink/show?code=XZ6msZu8BOExSFH4SxixhuxoyrmbRYGQQk Misc]
 
|[[User:Alsa|Alsa]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Surface Textures
 
|220K
 
|32
 
|128
 
|20.797K
 
|Custom (Photos)
 
|[http://www.mediafire.com/file/w3jujtm752hvdj1/Manga109Attempt.pth.zip/file Manga109Attempt]
 
|-
 
|[https://drive.google.com/file/d/1d89zvzC6BKf5EKapBtc4MH2DX5GClKzl/view Nickelback]
 
|[[User:BlackScout|BlackScout]]
 
|4
 
|[https://www.gnu.org/licenses/gpl-3.0.en.html GNU GPLv3]
 
|Photographs
 
|70k
 
|8
 
|128
 
|44.55k
 
|Wallpapers
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|-
 
|[https://drive.google.com/file/d/1IHp3Ibs3B8wSKZg2iGAVy4DP5bTSTOBb/view Fuzzy Box]
 
|BlueAmulet
 
|4
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|Anything
 
|197K
 
|16
 
|128
 
|9.962K
 
|CC0 Textures
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|-
 
|[https://drive.google.com/file/d/1KToK9mOz05wgxeMaWj9XFLOE4cnvo40D/view?usp=sharing Box]
 
|Jacob_ (buildist)
 
|4
 
|[https://www.gnu.org/licenses/gpl-3.0.en.html GNU GPLv3]
 
|Realistic
 
|390K
 
|8
 
|192
 
|11.577K
 
|Flickr2K+Div2K+OST
 
|PSNR model from same data
 
|-
 
|[https://drive.google.com/u/1/uc?id=10h8YXKKOQ61ANnwLjjHqXJdn4SbBuUku&export=download Lollypop]
 
|[[User:LyonHrt|LyonHrt]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Universal
 
|235k
 
|8
 
|128
 
|109.5K
 
|prerendered images
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|-
 
|[https://u.pcloud.link/publink/show?code=kZOhRakZOUkSXYB6at7ndvDxVJBGQLSaIfBk Superscale]
 
|[[User:Nmkd|nmkd]]
 
|4
 
|[http://www.wtfpl.net/ WTFPL]
 
|Clean Real-World Images
 
|134k
 
|1/2
 
|448
 
|900
 
|DIV2K Train + Valid
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|-
 
|[https://drive.google.com/file/d/1dGmhHUPmb3lO9buX_Bt2nq97Nk5MCTb4/view Ground]
 
|ZaphodBreeblebox
 
|4
 
|
 
|Ground Textures
 
|305K
 
|?
 
|128
 
|?
 
|Custom (Ground textures Google)
 
|?
 
|}
 
 
 
Box was meant to be an improvement on the RRDB_ESRGAN_x4 model ([https://imgsli.com/NDEwMw comparison]). It’s also trained on photos, but with a much larger dataset which was downscaled with linear interpolation (box filter) instead of bicubic.
 
 
 
Deoldify is intended for denoising and deblurring photos. All outputs are greyscale. This was initially a proof of concept. ''(Need clarification on specific subject matter.)''
 
 
 
The Ground model was trained on various pictures of '''stones, dirt and grass''' using Google’s image search.
 
 
 
The Misc model is trained on various pictures shoot by myself, including '''bricks, stone, dirt, grass, plants, wood, bark, metal''' and a few others.
 
 
 
Superscale by NMKD was made to rival the RealSR models and is one of the best upscalers for real-world content. It's trained on PNGs, so it doesn't handle compression artifacts well.
 
 
 
Lollypop was originally made for prerendered manga faces, using the patchgan discriminator. Combined with cutmixup, cx loss, and Frequency Separator, it handles prerendered, realistic faces, and even pixel art, with dedithering low palette images incredibly well.
 
 
 
====Characters and Faces====
 
 
 
For images of humans, creatures, faces, ...
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://drive.google.com/file/d/1HIBRKFs7s-XhpN1p7rwAWcMTHwFRqn3n/view Trixie]
 
|[[User:LyonHrt|LyonHrt]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Star Wars
 
|275K
 
|8
 
|192
 
|19.814K
 
|?
 
|None
 
|-
 
|[https://drive.google.com/open?id=19ICMKNuS4PbhmtA7Be9lUE6v8NHDDEzP Face Focus]
 
|[[User:LyonHrt|LyonHrt]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Face De-blur
 
|275K
 
|8
 
|192
 
|4.157K
 
|Custom (Faces)
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|-
 
|[https://de-next.owncube.com/index.php/s/YbAqNMMTtxrajF6 Face]
 
|[[User:Twittman|Twittman]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Face Upscaling
 
|250K
 
|10
 
|128
 
|3.765K
 
|Custom (Faces)
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|-
 
|[https://de-next.owncube.com/index.php/s/mZRG4HB3KdP2iP6 Face-Ality V1]
 
|[[User:Twittman|Twittman]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Face Upscaling
 
|310k
 
|10
 
|128
 
|13.3k
 
|Custom (Faces)
 
|4x_Faces_04_N_180000_G.pth
 
|-
 
|[https://drive.google.com/open?id=1OyOJIW224hBhb-aTCbuUQb0qzKmE4oH6 TGHQFace8x]
 
|[[User:Torrentguy_(tg)|Torrentguy]]
 
|8
 
|[https://www.gnu.org/licenses/gpl-3.0.en.html GNU GPLv3]
 
|Face Upscaling
 
|500k
 
|8
 
|128
 
|70k
 
|Flickr Cropped Faces
 
|[https://mega.nz/#!6kwQiCCS!v2uN8R44vVrlzmSqffGaCnzgogkPhhl67myJbuG45SA 8xESRGAN]
 
|}
 
 
 
Trixie was made to bring balance to the force… Also to upscale '''character textures for star wars games''', including the heroes, rebels, sith and imperial. Plus a few main aliens…Why called trixie? Because jar jars big adventure would be too long of a name…This also provides good upscale for '''face textures for general purpose as well as basic star wars''' textures.
 
 
 
The Face Focus modes was designed for '''slightly out of focus / blurred images of faces'''. It is aimed at '''faces''' / '''hair''', but it can help to '''improve other out of focused images too''' as always just try it.
 
 
 
===Specialized===
 
 
 
For Purposes that didn't fit anywhere else (for now).
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://e.pcloud.link/publink/show?code=XZGmsZo1yNrot19O4R5SrPVD1GWpRy1MKX Skyrim Armory]
 
|[[User:Alsa|Alsa]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Armor, Clothes and Weapons
 
|80K
 
|26
 
|128
 
|0.8K
 
|Skyrim Mod textures
 
|[http://www.mediafire.com/file/w3jujtm752hvdj1/Manga109Attempt.pth.zip/file Manga109Attempt]
 
|-
 
|[https://drive.google.com/file/d/1zC97RLUg5ejqb5-4Dj5q-SzdLPLiMozH/view Nickelfront]
 
|[[User:BlackScout|BlackScout]]
 
|4
 
|[https://www.gnu.org/licenses/gpl-3.0.en.html GNU GPLv3]
 
|Money, Coins
 
|14k
 
|8
 
|128
 
|1.176k
 
|Images of Coins
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|-
 
|[https://drive.google.com/file/d/1zlc8BWOcpCbSZpwf7ArMCqpb0Tf9DlZk/view Wolly]
 
|[[User:BlackScout|BlackScout]]
 
|2
 
|[https://www.gnu.org/licenses/gpl-3.0.en.html GNU GPLv3]
 
|Pixar Movies / Wall-E
 
|38k
 
|1
 
|256
 
|9.0k
 
|BD 1080p Wall-E Frames
 
|[https://mega.nz/#!vtgSWKQT!K7Asn2zKe4N70R2aV89KEMTKhH3aiyGAAiuQDJF09qs 2xESRGAN]
 
|-
 
|[https://1drv.ms/u/s!AiWox1lAWLoTg1bUeJouJNOFZ_Jj Fallout 4 Weapons]
 
|Bob
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Fallout Weapon Diffuse Textures
 
|120K
 
|13
 
|128
 
|532(OTF)
 
|Fallout 4 HDDLC Weapon Diffuse Textures
 
|[http://www.mediafire.com/file/w3jujtm752hvdj1/Manga109Attempt.pth.zip/file Manga109Attempt]
 
|-
 
|[https://1drv.ms/u/s!AiWox1lAWLoThANuXjxIR-hqA6os Fallout Weapons V2]
 
|Bob
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Video game textures, mostly metal rusty, clean or painted.
 
|180K
 
|13
 
|128
 
|1.999K(OTF)
 
|Fallout 4 HD DLC(Weapon,Armor/Clothes,Vehicle and Interior/Architecture textures to be exact)
 
|[https://1drv.ms/u/s!AiWox1lAWLoTg1bUeJouJNOFZ_Jj Fallout 4 Weapons]
 
|-
 
|[https://drive.google.com/file/d/1RjVWNUFtVyz4ykqu6Pm9wmO03pPEBGqg/view Skyrim Misc]
 
|Deorder
 
|4
 
|[https://creativecommons.org/publicdomain/zero/1.0/ CC0]
 
|Skyrim Diffuse Textures
 
|105K
 
|?
 
|128
 
|?
 
|Skyrim Diffuse Textures
 
|?
 
|-
 
|[https://drive.google.com/file/d/1Xe1JorGFJzPClR9tUuGmpuMyLhTA-d2T/view?usp=sharing PackCraft V4]
 
|Joey
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Upscaling Minecraft pack.png
 
|40K
 
|1
 
|512
 
|60K
 
|Alpha Minecraft
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|-
 
|[https://drive.google.com/file/d/1waC7a63nh5qkrLQe95sMcN4HRpaqhD9r/view Skyrim Wood]
 
|Laeris
 
|4
 
|
 
|Wood
 
|75K
 
|?
 
|?
 
|?
 
|?
 
|?
 
|-
 
|[https://drive.google.com/open?id=12fR-pWw6YL2ZbS6EDvrAeOVoQHWkWl5L Forest]
 
|[[User:LyonHrt|LyonHrt]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Wood / Leaves
 
|160K
 
|8
 
|192
 
|2.2K
 
|Custom(?)
 
|none
 
|-
 
|[https://drive.google.com/open?id=1xPzqzTG5L6y5I0U3W0oQjexwGF5zfG32 Map]
 
|[[User:LyonHrt|LyonHrt]]
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Map / Old Paper with text
 
|120K
 
|8
 
|192
 
|2.311K
 
|Custom(Scans)
 
|none
 
|-
 
|[https://mega.nz/folder/tFhl0L5b#CK4WLHYxjX7KFXU1PLNDuQ Morrowind Mixed]
 
|mkultra
 
|4
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Morrowind Mod Textures
 
|115K-130K
 
|16
 
|128
 
|56K
 
|Morrowind Mod Textures
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|}
 
 
 
The map model was trained on '''maps, old documents, papers and various styles of typefaces/fonts'''. Based on a dataset contributed by alsa. [https://imgsli.com/MzMwNw Sample]
 
 
 
The Forest model is focused on '''trees, leaves, bark and stone''' can be used for double upscaling for even more detail. [https://imgsli.com/MzI2Nw Sample]
 
 
 
The Armory model was trained with '''modded textures form Skyrim''', including '''Clothing, Armor and Weapons'''. ('''Leather, Canvas and Metal''' should all work - maybe too sharp so interpolate)
 
 
 
The wood model was trained for Skyrim by Laeris.
 
 
 
The Skyrim Diffuse models is supposed to be used with Skyrim’s diffuse textures. It is a bit too sharp so I recommend to interpolating with the |[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN] model or the mangaAttempt109 model, look in [https://drive.google.com/drive/folders/1m1sgHiSZkmI4NbiXeBEZasf94WPqEsNF Deorder’s Skyrim Model Google Drive] for an already interpolated version.
 
 
 
Fallout 4 weapons was trained using Fallout 4’s official HDarmor/weapon textures but could be used on other weapon and armor textures.
 
 
 
===Grayscale===
 
 
 
In digital photography, computer-generated imagery, and colorimetry, a grayscale or greyscale image is one in which the value of each pixel is a single sample representing only an amount of light, that is, it carries only intensity information. Grayscale images, a kind of black-and-white or gray monochrome, are composed exclusively of shades of gray. The contrast ranges from black at the weakest intensity to white at the strongest.[1]
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://drive.google.com/file/d/1trYs4AuC9s2JWAbryHdNyy5cgYV-V8cH/view Skyrim Alpha]
 
|Deorder
 
|4
 
|[https://creativecommons.org/publicdomain/zero/1.0/ CC0]
 
|Alpha Channel
 
|105K
 
|?
 
|128
 
|?
 
|Alpha Channels from Skyrim
 
|?
 
|-
 
|[https://drive.google.com/open?id=1-mxmDF1Dh-PnQqRz_PeCrvsTkHjYCbi3 Deoldify]
 
|TheAtheistGod
 
|4
 
|[https://creativecommons.org/publicdomain/zero/1.0/ CC0]
 
|B/W Photo Restoration
 
|790K
 
|16
 
|128
 
|2.931M
 
|[https://groups.csail.mit.edu/vision/datasets/ADE20K/ ADE20K], [https://data.vision.ee.ethz.ch/cvl/DIV2K/ DIV2K], [http://www.loc.gov/pictures/ Library of Congress]
 
|Falcon Fanart (?)
 
|}
 
 
 
Trained to upscale grayscale images, like specular or alpha etc.
 
 
 
==Artifact Removal==
 
 
 
The models in this section were made to remove compression artifacts in images and textures.
 
 
 
===Blur===
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://drive.google.com/file/d/1GoUY7t5kE0Ubi-NDvlMB8lUs-_u_f6tS/view?usp=sharing LADDIER1]
 
|Alexander Syring
 
|4
 
|[https://creativecommons.org/publicdomain/zero/1.0/ CC0]
 
|Remove noise, grain, box/lens/guassian blur
 
|282K
 
|8
 
|128
 
|12.7K
 
|Mostly images of Nirvana, some of other images
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|-
 
|[https://de-next.owncube.com/index.php/s/aAojXwLTPZto8rP Fatality DeBlur]
 
|[[User:Twittman|Twittman]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Deblurring
 
|270K
 
|1
 
|128
 
|41K
 
|Mix of anime, manga, and photos
 
|1x_DeJpeg_Fatality_01_175000_G.pth
 
|}
 
 
 
Fatality DeBlur is intended for general deblurring of images, and has some robustness against compression and noise.
 
 
 
LADDIER1 upscales images with noise, grain and different types of blur.
 
 
 
===Noise===
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://drive.google.com/open?id=1f5kIr3gG5GjbN_EM96CCB0nqu_2_56gh ISO denoise v1]
 
|[https://github.com/alpha0010/ Alpha]
 
|1
 
|[https://creativecommons.org/publicdomain/zero/1.0/ CC0]
 
|Remove high ISO noise
 
|372K
 
|1
 
|128
 
|170K
 
|Custom (DSLR photos)
 
|[https://mega.nz/#!D940CAKR!-BAx7tSj1CgeopGkaMW31VuPsC3hvNBL35TEVAj4LIo 1xESRGAN]
 
|-
 
|[https://drive.google.com/open?id=1HSUvnQm56_CmVAAMjTfVn_Hi6a4Bm2h0 ISO denoise v2]
 
|[https://github.com/alpha0010/ Alpha]
 
|1
 
|[https://creativecommons.org/publicdomain/zero/1.0/ CC0]
 
|Remove high ISO noise
 
|308K
 
|1
 
|128
 
|239K
 
|Custom (DSLR photos)
 
|[https://drive.google.com/open?id=1f5kIr3gG5GjbN_EM96CCB0nqu_2_56gh ISO denoise v1]
 
|-
 
|[https://drive.google.com/file/d/1hnWvQ2VzzxaTaar6ifErFJKrCmYXH8v0/view?usp=sharing NoiseToner]
 
|[[User:Mutin Choler|Mutin Choler]]
 
|1
 
|[http://www.wtfpl.net/ WTFPL]
 
|Remove high ISO noise
 
|105K
 
|4
 
|64
 
|800
 
|Custom (Photos)
 
|[https://mega.nz/#!D940CAKR!-BAx7tSj1CgeopGkaMW31VuPsC3hvNBL35TEVAj4LIo 1xESRGAN]
 
|}
 
High ISO is required to capture images at very high shutter speeds and/or in poor lighting. However, increasing ISO also results in noise that is challenging to remove without also losing fine details. To remove this noise without damaging details, one can apply the technique of median stacking: combine 10-100 photos of the exact same object to separate the signal (clear photo) from noise.
 
 
 
ISO denoise v1 is trained on a handcraft set of pairs between original noisy photos and median stacked counterparts. It tends to be conservative, avoiding modifying anything that does not closely resemble camera ISO noise. ISO denoise v2 uses a larger dataset, handling some cases better and some cases worse than v1. If results are unsatisfactory with one model, try the other model.
 
 
 
===JPG Compression===
 
 
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://e.pcloud.link/publink/show?code=XZvmsZo4qORoatgHXm4A1CooSUnuy5YD5X jpg (0-20%) alsa]
 
|[[User:Alsa|Alsa]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|JPG compressed Images
 
|178K
 
|2
 
|128
 
|6.23K
 
|Custom (Photos / Manga)
 
|[https://mega.nz/#!OzpFHYAD!0swA_J18bSygyVgMHac3P4kkBx6ZYUHNnL9qXhiu-9I JPG (20-40%)]
 
|-
 
|[https://e.pcloud.link/publink/show?code=XZimsZmfeBi2HJp9H00wvCRISpnuSRrjgy jpg (20-40%) alsa]
 
|[[User:Alsa|Alsa]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|JPG compressed Images
 
|141K
 
|2
 
|128
 
|6.23K
 
|Custom (Photos / Manga)
 
|[https://mega.nz/#!Kr5Q1C7C!yq4910SjvAyMS_40IhrOsShA21OXdP8rxfiMsEuYbW8 JPG (40-60%)]
 
|-
 
|[https://e.pcloud.link/publink/show?code=XZrmsZbOaINRer4xRn6UWBtGgRK5r7biWy jpg (40-60%) alsa]
 
|[[User:Alsa|Alsa]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|JPG compressed Images
 
|100K
 
|2
 
|128
 
|~6.5K
 
|Custom (Photos / Manga)
 
|[https://mega.nz/#!6j4RRQBD!3CX7XN9ZEAeWR3aXBAsDRpxSfKbM8zR1N9GXgbP38Ig JPG (60-80%)]
 
|-
 
|[https://e.pcloud.link/publink/show?code=XZcmsZASeN84hQXCLfhYzUogomG5EdkwPV jpg (60-80%) alsa]
 
|[[User:Alsa|Alsa]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|JPG compressed Images
 
|91K
 
|2
 
|128
 
|~6.5K
 
|Custom (Photos / Manga)
 
|[https://mega.nz/#!2uxwRYYT!2KgAsRac70jqYYXyOvQ5HSd4ipoqC0WGB4Md2Su_OM8 JPG (80-100%)]
 
|-
 
|[https://e.pcloud.link/publink/show?code=XZomsZwGOr2Jb4lLXGopIwu8UtNQqpuqyX jpg (80-100%) alsa]
 
|[[User:Alsa|Alsa]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|JPG compressed Images
 
|162K
 
|2
 
|128
 
|~6.5K
 
|Custom (Photos / Manga)
 
|[https://mega.nz/#!emAAQa6B!xItl6e7WIyq1ZbXXWtLaKn7M2it7j8Rx_-wvxDjWdQQ BC1 take 1]
 
|-
 
|[https://drive.google.com/file/d/1AjVnJCne4_JNYKGfzeogNxXo5zXMCGLe/view jpg (20-40%) BlueAmulet]
 
|[[User:BlueAmulet|BlueAmulet]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|JPG compressed Images
 
|535K
 
|1
 
|128
 
|52.789K
 
|Custom (CC0 Textures)
 
|[https://mega.nz/#!Kr5Q1C7C!yq4910SjvAyMS_40IhrOsShA21OXdP8rxfiMsEuYbW8 JPG (40-60%)]
 
|-
 
|[https://drive.google.com/file/d/1maYmC5yyzWCC42X5O0HeDuepsLFh7AV4/view jpg (0-20%) BlueAmulet]
 
|[[User:BlueAmulet|BlueAmulet]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|JPG compressed Images
 
|610K
 
|1
 
|128
 
|52.789K
 
|Custom (CC0 Textures)
 
|[https://mega.nz/#!OzpFHYAD!0swA_J18bSygyVgMHac3P4kkBx6ZYUHNnL9qXhiu-9I JPG (20-40%)]
 
|-
 
|[https://drive.google.com/file/d/1lgpo60okKC_9D6mzymHOvwdJDsOqxi7j/view jpg (40-60%) BlueAmulet]
 
|[[User:BlueAmulet|BlueAmulet]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|JPG compressed Images
 
|550K
 
|1
 
|128
 
|52.789K
 
|Custom (CC0 Textures)
 
|[https://mega.nz/#!6j4RRQBD!3CX7XN9ZEAeWR3aXBAsDRpxSfKbM8zR1N9GXgbP38Ig JPG (60-80%)]
 
|-
 
|[https://drive.google.com/file/d/1V2PTYzMMDIB7NNCp-ZXhIf26a0xHO2b2/view jpg (60-80%) BlueAmulet]
 
|[[User:BlueAmulet|BlueAmulet]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|JPG compressed Images
 
|545K
 
|1
 
|128
 
|52.789K
 
|Custom (CC0 Textures)
 
|[https://mega.nz/#!2uxwRYYT!2KgAsRac70jqYYXyOvQ5HSd4ipoqC0WGB4Md2Su_OM8 JPG (80-100%)]
 
|-
 
|[https://drive.google.com/file/d/17ocFMDz4jfYH73mT_4i1099JCGVnOHd2/view jpg (80-100%) BlueAmulet]
 
|[[User:BlueAmulet|BlueAmulet]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|JPG compressed Images
 
|550K
 
|1
 
|128
 
|52.789K
 
|Custom (CC0 Textures)
 
|[https://mega.nz/#!emAAQa6B!xItl6e7WIyq1ZbXXWtLaKn7M2it7j8Rx_-wvxDjWdQQ BC1 take 1]
 
|-
 
|[https://drive.google.com/file/d/17k72r01xTkeOzqsaYCN6ZqnAktKqTLyw/view?usp=sharing Jaypeg90]
 
|Jacob_
 
|4
 
|[https://www.gnu.org/licenses/gpl-3.0.en.html GNU GPLv3]
 
|JPG Photos 85-95% Quality
 
|66K
 
|24
 
|128
 
|3.440K * 2
 
|Flickr2K+Div2K, 85-95% JPEG Compressed, cubic or NN scaling
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|-
 
|[https://u.pcloud.link/publink/show?code=kZvfeFXZMyUSYScxAizUmTKjax0JQzTk9Gfk SBDV-DeJPEG-Lite]
 
|Joey
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|JPG compressed Images
 
|130k
 
|12
 
|64
 
|934
 
|SpongeBob, DIV2K
 
|1x_DIV2K-Lite_80k.pth
 
|-
 
|[https://u.pcloud.link/publink/show?code=kZuDPakZRh7Wrg6YSBkUdHPm3NsyJ5GtI570 Jaywreck]
 
|[[User:Nmkd|nmkd]]
 
|1
 
|[http://www.wtfpl.net/ WTFPL]
 
|JPG compressed Images
 
|190k
 
|1/2
 
|132
 
|900
 
|DIV2K Train + Test
 
|[https://mega.nz/#!D940CAKR!-BAx7tSj1CgeopGkaMW31VuPsC3hvNBL35TEVAj4LIo 1xESRGAN]
 
|-
 
|[https://u.pcloud.link/publink/show?code=kZ8FtFXZN0MakfJ5eWzEkGNOddw5pjp1paN7 Jaywreck3]
 
|[[User:Nmkd|nmkd]]
 
|1
 
|[http://www.wtfpl.net/ WTFPL]
 
|JPG compressed Images
 
|320k
 
|4
 
|112
 
|1800 (900 unique)
 
|DIV2K Train + Valid
 
|[https://mega.nz/#!D940CAKR!-BAx7tSj1CgeopGkaMW31VuPsC3hvNBL35TEVAj4LIo 1xESRGAN]
 
|-
 
|[https://de-next.owncube.com/index.php/s/jetoMQkKNLfRkYY JPG PlusULTRA]
 
|[[User:Twittman|Twittman]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|JPG compressed Images
 
|130K
 
|1
 
|?
 
|0.937K
 
|Custom (Manga)
 
|Failed Attempts
 
|-
 
|[https://de-next.owncube.com/index.php/s/w82HLrLWmWi4SQ5 DeJpeg Fatality PlusULTRA]
 
|[[User:Twittman|Twittman]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|JPG compressed Images
 
|200K
 
|1
 
|128
 
|26K
 
|Real life, Manga and digital text
 
|1x_DeJpeg_Fatality_01_200000_G.pth
 
|}
 
 
 
JPG gets compressed with a Quality Percentage between 0 and 100. So depending on how bad your JPEGs are compressed, choose the model of your choice. You can use [https://imagemagick.org/index.php ImageMagick] to guess the Quality percentage, but keep in mind that it might be wrong, since the image might have been re-saved.
 
 
 
JPG PlusULTRA, while not considered successful in its initial purpose, may find some use on certain unique images.
 
 
 
DeJpeg Fatality PlusULTRA is its burly strongman big brother. Anything and everything that is JPG will tremble before you!
 
 
 
Jaywreck by NMKD is trained to handle any amount of JPEG compression, even removing artifacts of quality 5 JPEGs. Its only weakness is a slight tendency to oversharpen certain textures. It's recommended to not use this model for very small images ( < 256px due to the large HR size it was trained on.
 
 
 
===DDS Files with BC1/DXT1, BC3/DXT5 Compression===
 
 
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://e.pcloud.link/publink/show?code=XZXbsZK03Y0UsmtR0W7wRMHWGvYkzgPII7 bc1 sharp 3.0]
 
|[[User:Alsa|Alsa]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|BC1 Compression
 
|400K
 
|2
 
|128
 
|28.985K
 
|Custom (just about everything)
 
|[https://mega.nz/#!r2hEAAAb!zdk-Ka6VCqVnKThtfJVYM0NnVBSUyeqsjYs-NgKjLkc BC1 take 2]
 
|-
 
|[https://drive.google.com/open?id=1LHplsPRqhmjR28jGgRlEekeP_bvx3nUC bc1 smooth 2.0]
 
|[[User:BlueAmulet|BlueAmulet]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|BC1 Compression
 
|1M
 
|20
 
|32
 
|4767
 
|Custom (? including normal maps)
 
|none (no interpolation)
 
|-
 
|[https://drive.google.com/file/d/1LcBy7G3Xovgo38WySTQd1UM5PrINmJu1/view?usp=sharing BCGoneV2]
 
|[[User:Mutin Choler|Mutin Choler]]
 
|1
 
|[http://www.wtfpl.net/ WTFPL]
 
|BC1 Compression
 
|225K
 
|1
 
|128
 
|250
 
|Custom (Photos)
 
|[https://mega.nz/#!D940CAKR!-BAx7tSj1CgeopGkaMW31VuPsC3hvNBL35TEVAj4LIo 1xESRGAN]
 
|-
 
|[https://drive.google.com/file/d/1XS0uBDDhIpZGROFeASlfQrrTuUKc8k9r/view?usp=sharing BCGoneV3]
 
|[[User:Mutin Choler|Mutin Choler]]
 
|1
 
|[http://www.wtfpl.net/ WTFPL]
 
|BC1 Compression
 
|115K
 
|4
 
|64
 
|1400
 
|Images from DIV2K and random anime drawings
 
|[https://mega.nz/#!D940CAKR!-BAx7tSj1CgeopGkaMW31VuPsC3hvNBL35TEVAj4LIo 1xESRGAN]
 
|}
 
 
 
BC1 (DXT1) compression is commonly used in dds textures, which are utilized in most PC games today, it allows to shrink the texture to 1/6 of the original size, reducing VRAM usage. There is also BC3 (DXT5) which uses BC1 compression for the color channels and leaves the alpha channel uncompressed, this one reduces the file size to 1/3 of the original. But this compression comes at a cost. The BC1 models are designed to reverse the damage done by the compression. This is a must if you want to upscale a dds file that uses either BC1 or BC3 compression. The free variant has more freedom when dealing with the images and should lead to better results. The restricted version only deals with perfect images compressed once (ideal case) so will perform worse in any other scenario, but it tries to preserve the original colors more. You can interpolate between restricted and unrestricted if you want.
 
 
 
===Cinepak, msvideo1 and Roq===
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://drive.google.com/open?id=1TMhjBAKVuDivz7FrAt-Qu7wufEJV6VUq Cinepak_alt]
 
|buildist
 
|1
 
|[https://www.gnu.org/licenses/gpl-3.0.en.html GNU GPLv3]
 
|Cinepak (only)
 
|240K
 
|1
 
|128
 
|2,640
 
|Flickr2K
 
|none (no interpolation)
 
|-
 
|[https://de-next.owncube.com/index.php/s/wKjLmYsq7M5JAmx Cinepak]
 
|[[User:Twittman|Twittman]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Cinepak, msvideo1 and Roq
 
|200K
 
|1
 
|128
 
|~8K
 
|Custom (Manga)
 
|none (no interpolation)
 
|}
 
 
 
The Cinepak model removes movie compressions artifacts from older video compression methods like Cinepak, msvideo1 and Roq. Cinepak_alt may work better for video frames with realistic content.
 
 
 
===Indeo Compression Artifacts===
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://drive.google.com/file/d/1-lfgZ_6vtS9DsWU-57ig8UJExH26XmCv/view?usp=sharing DeIndeo]
 
|Wild West Quest#7975
 
|4
 
|WTFPL
 
|Indeo Compression Artifacts
 
|130K
 
|8
 
|128
 
|3.500K
 
|Custom
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|}
 
 
 
===Color Reduction===
 
 
 
====Dithering / GIF====
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://e.pcloud.link/publink/show?code=XZ5bsZtd0ldzEF2UFLSYqIBNhqNH21slxX dithering]
 
|[[User:Alsa|Alsa]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Dithered Images
 
|127K
 
|2
 
|128
 
|4.7K
 
|Custom (Photos / Manga)
 
|[https://mega.nz/#!7io2gSQR!UB9u2k51daixTgC2H0LdOWzNlkDyIDHwxBX4BVY2J3k JPG (0-20%)]
 
|-
 
|[https://drive.google.com/file/d/1H4KQyhcknOoExjvDdsoxAgTBMO7JuJ3w/view FSDedither]
 
|buildist (Jacob_)
 
|4
 
|[https://www.gnu.org/licenses/gpl-3.0.en.html GNU GPLv3]
 
|Photos/realistic images with dithering
 
|55K
 
|16
 
|128
 
|~20K
 
|Flickr2K+DIV2K dithered with GIMP
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|-
 
|[https://buildism.net/files/4xFSDedither_Riven.pth FSDedither Riven]
 
|buildist (Jacob_)
 
|4
 
|[https://www.gnu.org/licenses/gpl-3.0.en.html GNU GPLv3]
 
|Photos/realistic images with dithering
 
|200K
 
|24
 
|128
 
|~20K
 
|Flickr2K+DIV2K with different types of dithering applied
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|-
 
|[https://buildism.net/files/4xFSDedither_Riven_Smooth.pth FSDedither Riven Smooth]
 
|buildist (Jacob_)
 
|4
 
|[https://www.gnu.org/licenses/gpl-3.0.en.html GNU GPLv3]
 
|Photos/realistic images with dithering
 
|191K
 
|24
 
|128
 
|~40K
 
|Flickr2K+DIV2K with different types of dithering applied
 
|4xFSDedither_Riven
 
|-
 
|[https://u.pcloud.link/publink/show?code=kZlhTVXZdnHwYDPRY9FwsAenVEDf2BKFfjoy DeGif]
 
|[[User:Nmkd|nmkd]]
 
|2
 
|[http://www.wtfpl.net/ WTFPL]
 
|GIF Frames
 
|160K
 
|1
 
|192
 
|1.8K
 
|DIV2K Dithered + DIV2K Color Reduced
 
|[https://mega.nz/#!vtgSWKQT!K7Asn2zKe4N70R2aV89KEMTKhH3aiyGAAiuQDJF09qs 2xESRGAN]
 
|-
 
|[https://drive.google.com/file/d/1Hxaa-uoccLInScJA9FOzfjAZBHtA90-0/view?usp=sharing DitherDeleter32]
 
|[[User:Mutin Choler|Mutin Choler]]
 
|1
 
|[http://www.wtfpl.net/ WTFPL]
 
|32 bit dedithering
 
|90K
 
|4
 
|64
 
|800
 
|Images from DIV2K
 
|[https://mega.nz/#!D940CAKR!-BAx7tSj1CgeopGkaMW31VuPsC3hvNBL35TEVAj4LIo 1xESRGAN]
 
|-
 
|[https://drive.google.com/file/d/1B20zbC1s66q5PYAQqg9j-8aVJiR2rwOv/view?usp=sharing DitherDeleter64]
 
|[[User:Mutin Choler|Mutin Choler]]
 
|1
 
|[http://www.wtfpl.net/ WTFPL]
 
|64 bit dedithering
 
|70K
 
|4
 
|64
 
|800
 
|Images from DIV2K
 
|[https://mega.nz/#!D940CAKR!-BAx7tSj1CgeopGkaMW31VuPsC3hvNBL35TEVAj4LIo 1xESRGAN]
 
|}
 
 
 
Dithering is an older compression method, where the amount of colors gets reduced, if your image has few colors and have a noise like pattern in them try a De-Dither model.
 
Ordered dithering is a less common form of dithering that results in distinctive checkerboard/crosshatch patterns, which are misinterpreted as texture by models not trained on it. It’s often used on GIFs because the pattern is stable between frames.
 
 
 
4sFSDedither is made for photos/realistic images, but worth trying on other images that have reduced colors and dithering along with fine details. Trained using the ESRGAN-FS code for better details compared to plain ESRGAN.
 
 
 
4xFSDedither_Riven is a remake of 4xFSDedither that is fine-tuned to upscale images from the game Riven, but should be better in general, particularly on ordered dithering. I adjusted the dataset to have a better variety of dithering parameters, and turned up the HFEN and pixel loss to get better details and color restoration with less noise. 4xFSDedither_Riven_Smooth adds images with nearest-neighbor downscaling to the dataset for a smoother (but possibly less detailed) result.
 
 
 
====Banding====
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://drive.google.com/file/d/1wQ4QxTaVAszCnDuRPGOtbLoWcJHh0GIm/view DeBandizer]
 
|[[User:BlackScout|BlackScout]]
 
|1
 
|[https://www.gnu.org/licenses/gpl-3.0.en.html CC BY-NC 4.0]
 
|Images with Banding
 
|34k
 
|1
 
|128
 
|changing ~900
 
|Custom (Wallpapers)
 
|old attempts based on [https://mega.nz/#!D940CAKR!-BAx7tSj1CgeopGkaMW31VuPsC3hvNBL35TEVAj4LIo 1xESRGAN]
 
|-
 
|[https://drive.google.com/file/d/1GNguVMV4E5t_4dn23w6UMXVKN6Ff9XPt/view?usp=sharing N64clean]
 
|[[User:BlueAmulet|BlueAmulet]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Cleaning N64 textures
 
|277k
 
|20
 
|48
 
|4034 (not tiled)
 
|CC0 Textures
 
|[https://mega.nz/#!D940CAKR!-BAx7tSj1CgeopGkaMW31VuPsC3hvNBL35TEVAj4LIo 1xESRGAN]
 
|}
 
 
 
Banding is like dithering, but instead of adding patterns of pixels to simulate missing colors, it just clips them off.
 
 
 
====Color Removal / Grayscale====
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://drive.google.com/file/d/1UgxnymTfLaVhY3aaDuBRnTZ9ehckKVKJ Colorizer]
 
|[[User:BlackScout|BlackScout]]
 
|1
 
|[https://www.gnu.org/licenses/gpl-3.0.en.html GNU GPLv3]
 
|Coloring B/W images
 
|22k
 
|1
 
|128
 
|4.5045k
 
|Custom (800 Wallpapers)
 
|[https://mega.nz/#!D940CAKR!-BAx7tSj1CgeopGkaMW31VuPsC3hvNBL35TEVAj4LIo 1xESRGAN]
 
|}
 
Just removes all colors and makes the image grayscale, the models here should restore some color to them, should also work for old black and white images.
 
 
 
===Over-Sharpening===
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://mega.nz/#!qRZD3SLL!RqP7cISIcoW5YakPya9CXEWSEHWqUSKdWLQSdYKGa14 DeSharpen]
 
|loinne
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Oversharpened Images
 
|310K
 
|1
 
|128
 
|~3K
 
|Custom (?)
 
|Failed Attempts
 
|}
 
 
 
The De-Sharpen model was made for rare particular cases when the image was destroyed by applying noise, i.e. game textures or any badly exported photos. If your image does not have any over-sharpening, it won’t hurt them, leaving as is. In theory, this model knows when to activate and when to skip, also can successfully remove artifacts if only some parts of the image are over-sharpened, for example in image consisting of several combined images, 1 of them with sharpen noise. It is made to remove sharpen noise, particularly made with Photoshop “sharpen” or “sharpen more” filters OR ImageMagick’s -sharpen directive with several varying parameters of Radius and Sigma, from subtle 0.3x0.5 to something extreme like 8x2, somewhere about that.
 
 
 
===Aliasing===
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://drive.google.com/open?id=11nMqVF9sUN2jRRagYc3wdWNX5acEy0MJ SS Anti Alias 9x]
 
|[[User:BlueAmulet|BlueAmulet]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc/4.0/ CC BY-NC 4.0]
 
|Anti aliasing
 
|125K~
 
|12
 
|32
 
|4096
 
|Custom (9x Supersampling AA)
 
|none (no interpolation)
 
|-
 
|[https://de-next.owncube.com/index.php/s/mmYCcK4H3rQQR5Y Anti Aliasing]
 
|[[User:Twittman|Twittman]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Images with pixelated edges
 
|200K
 
|1
 
|128
 
|0.656K
 
|Custom (?)
 
|none (no interpolation)
 
|}
 
 
 
AntiAliasing is for smoothing jagged edges in images and textures.
 
 
 
==Image Generation==
 
 
 
===Texture Maps===
 
 
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://drive.google.com/file/d/1uksv1uyyZjQcU4567OrfU0w2NECcG8dY/view normal generator]
 
|[[User:LyonHrt|LyonHrt]]
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Diffuse to Normal
 
|215K
 
|1
 
|128
 
|4.536K
 
|Custom (?)
 
|none (no interpolation)
 
|}
 
 
 
===DeTooning===
 
 
 
{| class="wikitable sortable"
 
|-
 
{{ModelDBColumns|Name|Author|Scale|License|Purpose|Iterations|Batch Size|HR Size|Dataset Size|Dataset|Pretrained Model}}
 
|-
 
|[https://drive.google.com/file/d/12vkDn35cuVnq2KYpkaA3XfCmk6zY_u4_/view?usp=sharing Double Detoon]
 
|Joey
 
|1
 
|[https://creativecommons.org/licenses/by-nc-sa/4.0/ CC BY-NC-SA 4.0]
 
|Detooning
 
|100K
 
|4
 
|64
 
|6.604K
 
|Reddit Images run through [https://github.com/SystemErrorWang/White-box-Cartoonization White-box-Cartoonization]
 
|[https://mega.nz/#!D940CAKR!-BAx7tSj1CgeopGkaMW31VuPsC3hvNBL35TEVAj4LIo 1xESRGAN]
 
|}
 
 
 
==Pretrained models for different scales==
 
{| class="wikitable sortable"
 
|-
 
!Name!!Author!!Scale!!License!!Dataset Size!!Dataset!!Pretrained Model
 
|-
 
|[https://mega.nz/#!D940CAKR!-BAx7tSj1CgeopGkaMW31VuPsC3hvNBL35TEVAj4LIo 1xESRGAN]
 
|[https://github.com/victorca25 victorca25]
 
|1
 
|Apache License 2.0
 
|65k
 
|Combination of DIV2K, Flickr2K and GOPRO
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|-
 
|[https://mega.nz/#!vtgSWKQT!K7Asn2zKe4N70R2aV89KEMTKhH3aiyGAAiuQDJF09qs 2xESRGAN]
 
|[https://github.com/victorca25 victorca25]
 
|2
 
|Apache License 2.0
 
|65k
 
|Combination of DIV2K, Flickr2K and GOPRO
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|-
 
|[https://mega.nz/#!uhAVEaAA!qfwIQ44Ba3rXMAb2-M3aM3zFBCwzmyQ64IO_0O-csJE 4xESRGAN]
 
|xinntao
 
|4
 
|Apache License 2.0
 
|65k
 
|Combination of DIV2K, Flickr2K and GOPRO
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|-
 
|[https://mega.nz/#!6kwQiCCS!v2uN8R44vVrlzmSqffGaCnzgogkPhhl67myJbuG45SA 8xESRGAN]
 
|[https://github.com/victorca25 victorca25]
 
|8
 
|Apache License 2.0
 
|65k
 
|Combination of DIV2K, Flickr2K and GOPRO
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|-
 
|[https://mega.nz/#!btwEXAoI!nNxWI89OdQCEkcRoPEFwj6GoFN8uzZyZFaCn1wbhzKY 16xESRGAN]
 
|[https://github.com/victorca25 victorca25]
 
|16
 
|Apache License 2.0
 
|65k
 
|Combination of DIV2K, Flickr2K and GOPRO
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|-
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 1xPSNR]
 
|[[User:BlueAmulet|BlueAmulet]]
 
|1
 
|Apache License 2.0
 
|65k
 
|Combination of DIV2K, Flickr2K and GOPRO
 
|none
 
|-
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 2xPSNR]
 
|[[User:BlueAmulet|BlueAmulet]]
 
|2
 
|Apache License 2.0
 
|65k
 
|Combination of DIV2K, Flickr2K and GOPRO
 
|none
 
|-
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 4xPSNR]
 
|xinntao
 
|4
 
|Apache License 2.0
 
|65k
 
|Combination of DIV2K, Flickr2K and GOPRO
 
|none
 
|-
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 8xPSNR]
 
|[[User:BlueAmulet|BlueAmulet]]
 
|8
 
|Apache License 2.0
 
|65k
 
|Combination of DIV2K, Flickr2K and GOPRO
 
|none
 
|-
 
|[https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL 16xPSNR]
 
|[[User:BlueAmulet|BlueAmulet]]
 
|16
 
|Apache License 2.0
 
|65k
 
|Combination of DIV2K, Flickr2K and GOPRO
 
|none
 
|-
 
|}
 
https://drive.google.com/drive/folders/1ldwajXL50uC7PCS63B4Wato6Dnk-svNL
 
 
 
These models were transformed from the original RRDB_ESRGAN_x4.pth model into the other scales, in order to be used as pretrained models for new models in those scales.
 
 
 
More information can be found [https://github.com/victorca25/BasicSR/wiki/Interpolation here].
 

Latest revision as of 04:21, 10 August 2023

Info-icon.png We have moved to https://openmodeldb.info/. This new site has a tag and search system, which will make finding the right models for you much easier! If you have any questions, ask here: https://discord.gg/cpAUpDK

If you need to look at the old Model Database anyway, look here