## Ani4K v2  

 Ani4K v2, as the successor to the original Ani4K, retains its predecessor's fantastic detail retention, depth of field preservation and faithfulness to the original source. As the name suggests, the model is targeted at modern anime, ranging from high-quality Bluray to crappy WEB releases, for upscaling either to 2K or 4K.

An **UltraCompact** version of the model is available on the [Github release](https://github.com/Sirosky/Upscale-Hub/releases/tag/Ani4K-v2). The UltraCompact version is faster, without a perceptual hit to quality in most cases.

📌 [More comparisons](https://slow.pics/c/1HykdyT5)

****

**FAQ**

- *How does v2 differ from v1?* I'm so glad you asked! A shortcoming of the v1 models is that the model really struggled on sources which were poorly mastered. This is unfortunately still very common even with modern anime. v2 is far more capable of dealing with such sources. 
- *How does Ani4K v2 differ from JaNai v3? Which one should I pick?* JaNai v3 is a fantastic model, and shares many of the fundamental training objectives behind Ani4K (DOF preservation, faithfulness, etc.). I'd say that the primary difference is one of training philosophy-- JaNai seeks to render the source as if it was originally mastered in 4K, whereas Ani4K seeks to produce an upscale that is as close as possible to the source (while cleaning up any issues). Long story short, test both models and see which you prefer.
- *What model versions are there?* Ani4K comes in Compact and UltraCompact flavors. Compact is of course a standard option. UltraCompact provides a noticeable performance uplift, without too much impact on quality. I ultimately did not train a SuperUltraCompact variant as I felt the hit to the model quality was far too significant.

Ani4VK v2 Compact

Purpose: Anime

This is a successor to AnimeSharpV3 based on RCAN instead of ESRGAN. It outperforms both versions of AnimeSharpV3 in every capacity. It's sharper, retains *even more* detail, and has very few artifacts. It is __extremely faithful__ to the input image, even with heavily compressed inputs. 

To use this model, you must update to the [**latest chaiNNer nightly build**](https://github.com/chaiNNer-org/chaiNNer-nightly/releases)

The `2x-AnimeSharpV4_Fast_RCAN_PU` model is trained on RCAN PixelUnshuffle. This is much faster, but comes at the cost of quality. I believe the model is ~95% the quality of the full V4 RCAN model, but ~6x faster in Pytorch and ~4x faster in TensorRT. This model is ideal for video processing, and as such was trained to handle MPEG2 & H264 compression.

__Comparisons:__ https://slow.pics/c/63Qu8HTN
https://slow.pics/c/DBJPDJM9
![1736292155 679079](https://github.com/user-attachments/assets/875be696-0948-4de2-9d03-85a581891a49)

AnimeSharpV4

AnimeSharpV4_Fast_RCAN_PU

Purpose: 2x anime upscaler which tries to preserve dof

2x anime upscaler, a non-otf version of the HFA2kCompact model which tries to preserve dof by not deblurring.

Original Link: https://drive.google.com/drive/folders/1RZbALUTCf4DYDTVMBR74b8Nt-7m6TEyk?usp=drive_link

HFA2k compact multijpg

Upscaling frames from Irodori anime (namely kemono friends) from 540p (the source render resolution) to 1080p, low resolution flat shaded art, de-JPEG of the aforementioned

KemonoScale v2

SPAN model for live action film and digital video. The main goal is to fix/reduce common video quality problems while maintaining fidelity. I tried the existing video-focused models and they all denoise or cause colour shifts so I decided to train my own.

The model is trained with compression (JPEG, MPEG-4 ASP, H264, VP9, H265), chroma subsampling, blurriness from multiple scaling, uneven horizontal and vertical resolution, oversharpening halos, bad deinterlacing jaggies, and onscreen text. It is not trained to remove noise at all so it preserves details in the source well. To prevent colour/brightness shifts, I used consistency loss in neosr. I had to modify consistency loss to use a stronger blur so it doesn't interfere with the halo removal.

Limitations:
1. The model has limited ability to see details through heavy grain, but light to moderate grain is fine.
2. The model still does not handle bad deinterlacing perfectly, especially if the source is vertically resized. Fixing bad deinterlacing is not the main goal so it is what it is. Sources that are line-doubled throughout should be descaled back to half height first for best results.
3. The model sometimes oversharpens a little. This is probably because the training data has some oversharpened images.
4. This model generally cannot handle VHS degradation.

More comparisons: https://slow.pics/c/DtDN7gaq

The training config and image degradation scripts used to create training data can be found in https://github.com/jcj83429/upscaling/tree/9332e7d5b07747ff347e5abdc43f8144364de9f7/2xLiveActionV1_SPAN

2xLiveActionV1_SPAN

Upscale Anime while keeping as much of the original detail as possible. Isn't among the sharpest models and if the anime is somewhat blurry it will retain that detail but add more pixels to give a smoother edge. Doesn't change the color very much and may make some scene's very slightly darker.

Pretrained model: Pony Models

Loyaldk Giroro

Loyaldk Keroro

Loyaldk Kororo

Upscale MLP episodes. Liter version. Able to handle compression better compared to V1.0 and no longer creates halo's/rainbows. Good for Vector 2D art however converts detail to blobs.

Loyaldk LitePony V2.0

Made to restore screentones and remove compression artifacts in manga images with widths between 650 and 900 pixels (can be more than 900).

2x Manga Ora

2x compact fast universal upscaler pair trained with jpg degradation (down to 40) and multiscale (down_up, bicubic, bilinear, box, nearest, lanczos).

2x-NomosUni_compact_multijpg

Purpose: 2x fast universal upscaler with medium degradation handling (jpg compression, noise, blur)

2x compact fast universal upscaler with medium degradation handling using the Real-ESRGAN training pipeline, based off 2xNomosUni_compact_otf_strong. Handles jpg compression, some noise, and some blur (so dejpgs, denoises and deblurs).

Original Link: https://drive.google.com/drive/folders/1YdOPyb0Ht9YjoRZILf87hSocyAx43CLc

Examples:
[RealPhoto](https://imgsli.com/MjMyNTUy)
[Noisy](https://imgsli.com/MjMyNTUz)

NomosUni compact otf medium

Purpose: 2x universal upscaler

2x esrgan universal upscaler, pair trained with jpg degradation (down to 40) and multiscale (down_up, bicubic, bilinear, box, nearest, lanczos).

Original Link: https://drive.google.com/drive/folders/12zKVS74mz0NtBKGlkx_r0ytUVoeDiTX3?usp=drive_link

Try it out in [this huggingface space](https://huggingface.co/spaces/Phips/upscale) I made (since the spaces uses cpu, testing locally on your gpu would be faster)

Slow Pics crops examples:
[ani](https://slow.pics/c/LLP9p89w)
[face](https://slow.pics/c/4IsxVMAB)
[dof_wheat](https://slow.pics/c/zj4q6Mvi)

NomosUni esrgan multijpg

2x span fast universal upscaler pair trained with jpg degradation (down to 40) and multiscale (down_up, bicubic, bilinear, box, nearest, lanczos).

2x-NomosUni_span_multijpg

Purpose: 2x fast universal DoF preserving upscaler

2x span fast universal DoF preserving upscaler, pair trained with jpg degradation (down to 40) and multiscale (down_up, bicubic, bilinear, box, nearest, lanczos).

Original Link: https://drive.google.com/drive/folders/1_P56B3k1XrBsa-f698zPHAuLaqQ8sSWF

Examples:
[dofwheat](https://slow.pics/c/k0U59uJC)
[digitalart](https://slow.pics/c/yyY0xMAn)  
[face](https://slow.pics/c/MdV3Lr7N)

NomosUni span multijpg ldl

Purpose: Compression Removal, Noise Reduction, Line Correction, MPEG2 / LD Artifact Removal

This is my first model release. The model was inspired from a personal project which I have been pursuing for some time now, which this model aims to solve. This model will upscale low resolution hand drawn animation from 1970s to 2000. The colors are retained with effective noise control. Details and Textures are maintained to a good degree considering animations. Color Spills are also corrected depending on the colors. Shades of white and yellows have been difficult. It also makes the lines slightly sharper and thinner. This could be a plus depending on your source. The model is also temporally stable across my tests with little observable issues. 

**Showcase:** 
Images - https://imgsli.com/MjYwNzY1/12/13
Video Sample -  https://t.ly/Jp7-w vs Upscale - https://t.ly/PdsKs

2x Pooh V4

In addition to removing the vertical blur, the model upscales, sharpens and will remove MPEG-2 artifacting and a small amount of rainbowing and dot crawl. Another series afflicted with the vertical blur is Avatar the Last Airbender, which can be repaired by this model. The video fed into the model MUST be 540 vertical for the deblur to work properly.

SwatKats

Purpose: Upscale text in very low quality to normal quality.

The upscale model is specifically designed to enhance lower-quality text images, improving their clarity and readability by upscaling them by 2x. It excels at processing moderately sized text, effectively transforming it into high-quality, legible scans. However, the model may encounter challenges when dealing with very small text, as its performance is optimized for text of a certain minimum size. For best results, input images should contain text that is not excessively small.

Text2HD v.1

[SwinIR: Image Restoration Using Swin Transformer](https://github.com/JingyunLiang/SwinIR)


Image restoration is a long-standing low-level vision problem that aims to restore high-quality images from low-quality images (e.g., downscaled, noisy and compressed images). While state-of-the-art image restoration methods are based on convolutional neural networks, few attempts have been made with Transformers which show impressive performance on high-level vision tasks. In this paper, we propose a strong baseline model SwinIR for image restoration based on the Swin Transformer. SwinIR consists of three parts: shallow feature extraction, deep feature extraction and high-quality image reconstruction. In particular, the deep feature extraction module is composed of several residual Swin Transformer blocks (RSTB), each of which has several Swin Transformer layers together with a residual connection. We conduct experiments on three representative tasks: image super-resolution (including classical, lightweight and real-world image super-resolution), image denoising (including grayscale and color image denoising) and JPEG compression artifact reduction. Experimental results demonstrate that SwinIR outperforms state-of-the-art methods on different tasks by up to 0.14~0.45dB, while the total number of parameters can be reduced by up to 67%.

SwinIR-M-x2 (classicalSR-DF2K-s64w8)

SwinIR-M-x2 (classicalSR-DIV2K-s64w8)

SwinIR-M-x2 (lightweightSR-DIV2K-s64w8)

[Real-ESRGAN](https://github.com/xinntao/Real-ESRGAN) aims at developing Practical Algorithms for General Image/Video Restoration.

Real-ESRGAN aims at developing Practical Algorithms for General Image/Video Restoration.
We extend the powerful ESRGAN to a practical restoration application (namely, Real-ESRGAN), which is trained with pure synthetic data.

RealESRGAN_x2Plus

4x photo upscaler on the SSDIR_Sharp dataset, trained with otf, using the same settings as Real_HAT_GAN.  
These otf values are very high in my opinion. This was an experiment if I used the same settings as Real_HAT_GAN (therefore the name).   
This model will denoise very strongly and smooth out a lot (so a lot of details will be lost. But maybe this effect might be beneficial to someone)

4xReal_SSDIR_DAT_GAN

Vast improvement over v1 in low frequency detail; moiré and artifacting reduced significantly and less random noise from JPEG artefacts in the input. Also now only works on 1 channel images, so it runs slightly faster on average and resulting images are much smaller but it might not work on some ESRGAN implementations, I personally recommend using chaiNNer. v1 may still be better in some edge cases.

There's also a supplementary 1x model that denoises very low quality LRs and smooths halftones so the image works better with the 4x model. Only trained it to help build the dataset and it's useless for already decent-ish LRs but may help you in some situations.

Architecture	DAT
Scale	2x
Size	s64 \| 8x3260dim3conv
Color Mode	RGB
License	CC-BY-NC-SA-4.0 Private use Distribution Modifications Credit required Same License State Changes No Liability & Warranty Disclaimer
Date	2023-12-08
Dataset size	4000
Training iterations	450000
Training HR size	128
Training OTF	No

2x Manga Ora

Similar Models