# [MAXIM](https://arxiv.org/abs/2201.02973): Multi-Axis MLP for Image Processing (CVPR 2022 Oral)
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/maxim-multi-axis-mlp-for-image-processing/deblurring-on-hide-trained-on-gopro)](https://paperswithcode.com/sota/deblurring-on-hide-trained-on-gopro?p=maxim-multi-axis-mlp-for-image-processing)
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/maxim-multi-axis-mlp-for-image-processing/deblurring-on-gopro)](https://paperswithcode.com/sota/deblurring-on-gopro?p=maxim-multi-axis-mlp-for-image-processing)
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/maxim-multi-axis-mlp-for-image-processing/deblurring-on-realblur-j-1)](https://paperswithcode.com/sota/deblurring-on-realblur-j-1?p=maxim-multi-axis-mlp-for-image-processing)
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/maxim-multi-axis-mlp-for-image-processing/deblurring-on-realblur-r)](https://paperswithcode.com/sota/deblurring-on-realblur-r?p=maxim-multi-axis-mlp-for-image-processing)
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/maxim-multi-axis-mlp-for-image-processing/deblurring-on-realblur-j-trained-on-gopro)](https://paperswithcode.com/sota/deblurring-on-realblur-j-trained-on-gopro?p=maxim-multi-axis-mlp-for-image-processing)
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/maxim-multi-axis-mlp-for-image-processing/deblurring-on-realblur-r-trained-on-gopro)](https://paperswithcode.com/sota/deblurring-on-realblur-r-trained-on-gopro?p=maxim-multi-axis-mlp-for-image-processing)
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/maxim-multi-axis-mlp-for-image-processing/low-light-image-enhancement-on-lol)](https://paperswithcode.com/sota/low-light-image-enhancement-on-lol?p=maxim-multi-axis-mlp-for-image-processing)
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/maxim-multi-axis-mlp-for-image-processing/photo-retouching-on-mit-adobe-5k)](https://paperswithcode.com/sota/photo-retouching-on-mit-adobe-5k?p=maxim-multi-axis-mlp-for-image-processing)
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/maxim-multi-axis-mlp-for-image-processing/single-image-deraining-on-rain100h)](https://paperswithcode.com/sota/single-image-deraining-on-rain100h?p=maxim-multi-axis-mlp-for-image-processing)
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/maxim-multi-axis-mlp-for-image-processing/single-image-deraining-on-rain100l)](https://paperswithcode.com/sota/single-image-deraining-on-rain100l?p=maxim-multi-axis-mlp-for-image-processing)
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/maxim-multi-axis-mlp-for-image-processing/single-image-deraining-on-test100)](https://paperswithcode.com/sota/single-image-deraining-on-test100?p=maxim-multi-axis-mlp-for-image-processing)
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/maxim-multi-axis-mlp-for-image-processing/single-image-deraining-on-test2800)](https://paperswithcode.com/sota/single-image-deraining-on-test2800?p=maxim-multi-axis-mlp-for-image-processing)
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/maxim-multi-axis-mlp-for-image-processing/single-image-deraining-on-test1200)](https://paperswithcode.com/sota/single-image-deraining-on-test1200?p=maxim-multi-axis-mlp-for-image-processing)
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/maxim-multi-axis-mlp-for-image-processing/image-denoising-on-sidd)](https://paperswithcode.com/sota/image-denoising-on-sidd?p=maxim-multi-axis-mlp-for-image-processing)
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/maxim-multi-axis-mlp-for-image-processing/image-denoising-on-dnd)](https://paperswithcode.com/sota/image-denoising-on-dnd?p=maxim-multi-axis-mlp-for-image-processing)
This repo is a PyTorch re-implementation of [**CVPR 2022 Oral**] paper: ["**MAXIM**: Multi-Axis MLP for Image Processing"](https://arxiv.org/abs/2201.02973) by [Zhengzhong Tu](https://www.linkedin.com/in/vztu/), [Hossein Talebi](https://scholar.google.com/citations?hl=en&user=UOX9BigAAAAJ), [Han Zhang](https://sites.google.com/view/hanzhang), [Feng Yang](https://sites.google.com/view/feng-yang), [Peyman Milanfar](https://sites.google.com/view/milanfarhome/), [Alan Bovik](https://www.ece.utexas.edu/people/faculty/alan-bovik), and [Yinxiao Li](https://scholar.google.com/citations?user=kZsIU74AAAAJ&hl=en)
Google Research, University of Texas at Austin
*__Disclaimer__: This repo is currently working in progress. No timelines are guaranteed.*
#### News
- **April 12, 2022:** Initialize PyTorch repo for MAXIM.
- **March 29, 2022:** The official JAX code and models have been released at [[google-research/maxim]](https://github.com/google-research/maxim)
- **March 29, 2022:** MAXIM is selected for an **ORAL presentation** at CVPR 2022 :tada:
- **March 3, 2022:** Paper accepted at CVPR 2022.
<hr />
> **Abstract:** *Recent progress on Transformers and multi-layer perceptron (MLP) models provide new network architectural designs for computer vision tasks. Although these models proved to be effective in many vision tasks such as image recognition, there remain challenges in adapting them for low-level vision. The inflexibility to support high-resolution images and limitations of local attention are perhaps the main bottlenecks. In this work, we present a multi-axis MLP based architecture called MAXIM, that can serve as an efficient and flexible general-purpose vision backbone for image processing tasks. MAXIM uses a UNet-shaped hierarchical structure and supports long-range interactions enabled by spatially-gated MLPs. Specifically, MAXIM contains two MLP-based building blocks: a multi-axis gated MLP that allows for efficient and scalable spatial mixing of local and global visual cues, and a cross-gating block, an alternative to cross-attention, which accounts for cross-feature conditioning. Both these modules are exclusively based on MLPs, but also benefit from being both global and `fully-convolutional', two properties that are desirable for image processing. Our extensive experimental results show that the proposed MAXIM model achieves state-of-the-art performance on more than ten benchmarks across a range of image processing tasks, including denoising, deblurring, deraining, dehazing, and enhancement while requiring fewer or comparable numbers of parameters and FLOPs than competitive models.*
<hr />
## Architecture
![Model overview](images/overview.png)
## Installation
TBD
<!-- Install dependencies:
```
pip install -r requirements.txt
``` -->
## Results and Pre-trained models
TBD
<!-- We provide all the pre-trained models and visual results.
| Task | Dataset | PSRN | SSIM | Model | #params | FLOPs | ckpt | outputs |
|:---:|:---:|:---:|:---:| :---:|:---:|:---:|:---:|:---:|
| Denoising | SIDD | 39.96 | 0.960 | MAXIM-3S | 22.2M | 339G | [ckpt](https://console.cloud.google.com/storage/browser/gresearch/maxim/ckpt/Denoising/SIDD/) | [images](https://console.cloud.google.com/storage/browser/gresearch/maxim/results/Denoising/SIDD/) |
| Denoising | DND | 39.84 | 0.954 | MAXIM-3S | 22.2M | 339G | [ckpt](https://console.cloud.google.com/storage/browser/gresearch/maxim/ckpt/Denoising/SIDD/) | [images](https://console.cloud.google.com/storage/browser/gresearch/maxim/results/Denoising/DND/) |
| Deblurring | GoPro | 32.86 | 0.961 | MAXIM-3S | 22.2M | 339G | [ckpt](https://console.cloud.google.com/storage/browser/gresearch/maxim/ckpt/Deblurring/GoPro) | [images](https://console.cloud.google.com/storage/browser/gresearch/maxim/results/Deblurring/GoPro/) |
| Deblurring | HIDE | 32.83 | 0.956 | MAXIM-3S | 22.2M | 339G | [ckpt](https://console.
没有合适的资源?快使用搜索试试~ 我知道了~
[CVPR2022Oral]PyTorch重新实现“MAXIM:用于图像处理的多轴MLP”,带有训练代码.zip
共135个文件
py:67个
png:25个
md:12个
1.该资源内容由用户上传,如若侵权请联系客服进行举报
2.虚拟产品一经售出概不退款(资源遇到问题,请及时私信上传者)
2.虚拟产品一经售出概不退款(资源遇到问题,请及时私信上传者)
版权申诉
0 下载量 73 浏览量
2023-04-28
13:46:04
上传
评论
收藏 10.79MB ZIP 举报
温馨提示
[CVPR2022Oral]PyTorch重新实现“MAXIM:用于图像处理的多轴MLP”,带有训练代码.zip
资源推荐
资源详情
资源评论
收起资源包目录
[CVPR2022Oral]PyTorch重新实现“MAXIM:用于图像处理的多轴MLP”,带有训练代码.zip (135个子文件)
setup.cfg 558B
.gitignore 73B
0.jpg 598KB
0.jpg 346KB
LICENSE 11KB
evaluate_PSNR_SSIM.m 8KB
evaluate_gopro_hide.m 1KB
evaluate_sidd.m 708B
README.md 15KB
README.md 5KB
README.md 2KB
README.md 1KB
INSTALL.md 1KB
README.md 1KB
README.md 1KB
README.md 880B
README.md 719B
README.md 134B
README.md 133B
README.md 133B
niqe_pris_params.npz 12KB
1fromGOPR1096.MP4.png 948KB
109fromGOPR1096.MP4.png 938KB
1fromGOPR0950.png 904KB
110fromGOPR1087.MP4.png 844KB
55.png 588KB
a4541-DSC_0040-2.png 578KB
55.png 526KB
overview.png 385KB
1.png 338KB
748.png 327KB
111.png 311KB
1444_10.png 293KB
1.png 284KB
0010_0.95_0.16.png 269KB
1.png 268KB
0014_0.8_0.12.png 257KB
0048_0.9_0.2.png 232KB
15.png 222KB
1440_10.png 185KB
0039_04.png 166KB
15.png 164KB
0003_30.png 161KB
0003_0.8_0.2.png 161KB
0011_23.png 156KB
0013_19.png 134KB
Maxim_arch.py 81KB
maxim.py 80KB
maxim_torch.py 80KB
base_model.py 14KB
data_util.py 14KB
paired_image_dataset.py 14KB
matlab_functions.py 13KB
video_test_dataset.py 12KB
image_restoration_model.py 12KB
train.py 12KB
restormer_arch.py 11KB
psnr_ssim.py 10KB
reds_dataset.py 10KB
face_util.py 9KB
transforms.py 9KB
arch_util.py 9KB
niqe.py 8KB
lr_scheduler.py 8KB
demo.py 8KB
img_util.py 7KB
lmdb_util.py 7KB
file_client.py 6KB
flow_util.py 6KB
logger.py 6KB
download_data.py 6KB
misc.py 6KB
setup.py 5KB
create_lmdb.py 5KB
__init__.py 5KB
vimeo90k_dataset.py 5KB
losses.py 4KB
jax2torch.py 4KB
test_gaussian_gray_denoising.py 4KB
test_gaussian_color_denoising.py 4KB
evaluate_realblur.py 4KB
download_data.py 4KB
generate_patches_gopro.py 4KB
options.py 4KB
bundle_submissions.py 4KB
fid.py 3KB
test_real_denoising_dnd.py 3KB
prefetch_dataloader.py 3KB
utils.py 3KB
utils.py 3KB
utils.py 3KB
test.py 3KB
test.py 3KB
loss_util.py 3KB
test_real_denoising_sidd.py 3KB
dist_util.py 3KB
single_image_dataset.py 2KB
evaluate_gaussian_gray_denoising.py 2KB
evaluate_gaussian_color_denoising.py 2KB
ffhq_dataset.py 2KB
共 135 条
- 1
- 2
资源评论
快撑死的鱼
- 粉丝: 1w+
- 资源: 9154
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功