update README.md

README.md

@@ -10,12 +10,22 @@
 
 This official repository contains the source code, prediction results, and evaluation toolbox of paper 'CamoFormer: Masked Separable Attention for Camouflaged Object Detection'. The technical report could be found at [arXiv](https://arxiv.org/abs/). 
 The whole benchmark results can be found at [One Drive](https://mailnankaieducn-my.sharepoint.com/:f:/g/personal/bowenyin_mail_nankai_edu_cn/EmB36EZb_fdMvWGgKx2EalgBuQnj8AFifyR-ip7Jtkfwqg?e=nu6DJz), [Baidu Netdisk](https://pan.baidu.com/s/1k5CxYzcgizzJ4sRdAxBNlA?pwd=srtf), or [Google Drive](https://drive.google.com/drive/folders/1gsCeYtS9cwsMpTHQzkx81n4jsRK4LYdf?usp=sharing).
-Code will be released soon.
+
+<p align="center">
+    <img src="figs/CamoFormer.png" width="600"  width="1200"/> <br />
+    <em> 
+    Figure 1: Overall architecture of our CamoFormer model. First, a pretrained Transformer-based backbone is utilized to extract multi-scale features of the input image. Then, the features from the last three stages are aggregated to generate the coarse prediction. Next, the
+progressive refinement decoder equipped with masked separable attention (MSA) is applied to gradually polish the prediction results. All
+the predictions generated by our CamoFormer are supervised by the ground truth (GT).
+    </em>
+</p>
+
+
 
 
 ## 1. :fire: NEWS :fire:
 
-- [2022/12/08] Releasing the whole COD benchmarking results (21 models).
+- [2022/12/09] Releasing the codebase of CamoFormer and the whole COD benchmarking results (21 models).
 - [2022/12/08] Creating repository.
 
 > We invite all to contribute in making it more acessible and useful. If you have any questions about our work, feel free to contact me via e-mail (bowenyin@mail.nankai.edu.cn). If you are using our code and evaluation toolbox for your research, please cite this paper ([BibTeX]()).
@@ -24,7 +34,27 @@ Code will be released soon.
 
 **0. Install**
 
-You could refer to [here](https://github.com/HVision-NKU/CamoFormer/blob/main/figs/Install.md).
+```
+conda create --name CamoFormer python=3.8.5
+conda activate CamoFormer
+conda install pytorch==1.12.1 torchvision==0.13.1 cudatoolkit=11.3 -c pytorch
+pip install opencv-python
+conda install tensorboard
+conda install tensorboardX
+pip install timm
+pip install matplotlib
+pip install scipy
+pip install einops
+
+Please also install [apex](https://github.com/NVIDIA/apex).
+git clone https://github.com/NVIDIA/apex
+cd apex
+pip install -v --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" ./
+or
+Apex also support Python-only build (required with Pytorch 0.4):
+pip install -v --no-cache-dir ./
+```
+
 
 **1. Download Datasets and Checkpoints.**
 
@@ -55,16 +85,7 @@ bash test_eval.sh
 
 ## 3. Proposed CamoFormer
 
-### 3.1. Overview
-
-<p align="center">
-    <img src="figs/CamoFormer.png" width="600"  width="1200"/> <br />
-    <em> 
-    Figure 1: Overall architecture of our CamoFormer model. First, a pretrained Transformer-based backbone is utilized to extract multi-scale features of the input image. Then, the features from the last three stages are aggregated to generate the coarse prediction. Next, the
-progressive refinement decoder equipped with masked separable attention (MSA) is applied to gradually polish the prediction results. All
-the predictions generated by our CamoFormer are supervised by the ground truth (GT).
-    </em>
-</p>
+### 3.1. The F-TA in MSA:
 
 
 <p align="center">