From 3d1c62c8d18bbf2b5437a4a002d7a68d62d0a0ae Mon Sep 17 00:00:00 2001
From: HantingChen <40243544+HantingChen@users.noreply.github.com>
Date: Sun, 27 Jun 2021 21:07:03 +0800
Subject: [PATCH] Update readme.md

---
 readme.md | 46 ++++++++++++++++++++++++++++++++++++++--------
 1 file changed, 38 insertions(+), 8 deletions(-)

diff --git a/readme.md b/readme.md
index 4614678..faed736 100644
--- a/readme.md
+++ b/readme.md
@@ -16,26 +16,52 @@ We study the low-level computer vision task (such as denoising, super-resolution
 - [Pretrained weights for super-resolution X4](https://www.mindspore.cn/resources/hub/details?noah-cvlab/gpu/1.1/ipt_v1.0_Set14_SR_x4)
 
 
+## Requirements
 
-## News
+- python 3
+- pytorch >= 1.4.0
+- torchvision
 
-- Pytorch pre-trained model will be released in June! 
 
+## Dataset
 
+The benchmark datasets can be downloaded as follows:
 
-## Requirements
+For super-resolution:
 
-- python 3
-- pytorch >= 1.4.0
-- torchvision
+ Set5,
+[Set14](https://sites.google.com/site/romanzeyde/research-interests),
+[B100](https://www2.eecs.berkeley.edu/Research/Projects/CS/vision/bsds/),
+Urban100.
+
+For denoising:
 
+[CBSD68](https://www2.eecs.berkeley.edu/Research/Projects/CS/vision/bsds/).
 
+For deraining:
 
-## Usage
+[Rain100L](https://www.icst.pku.edu.cn/struct/Projects/joint_rain_removal.html)
 
-Coming soon!
+The result images are converted into YCbCr color space. The PSNR is evaluated on the Y channel only.
 
+## Script Description
 
+This is the inference script of IPT, you can following steps to finish the test of image processing tasks, like SR, denoise and derain, via the corresponding pretrained models.
+
+### Script Parameter
+
+For details about hyperparameters, see option.py.
+
+## Evaluation
+
+### Evaluation Process
+
+> Inference example:
+> For SR x4:
+
+```bash
+python main.py --dir_data $DATA_PATH --data_test $DATA_TEST --test_only --ext img --pth_path $MODEL --task_id $TASK_ID --scale $SCALE
+```
 
 ## Results
 
@@ -126,3 +152,7 @@ Coming soon!
 ```
 
 
+## Acknowledgement
+
+* Main code from [EDSR-PyTorch](https://github.com/sanghyun-son/EDSR-PyTorch)  
+* Transformer code from [detr](https://github.com/facebookresearch/detr)