diff --git a/README.md b/README.md
index ea31033..21e625a 100644
--- a/README.md
+++ b/README.md
@@ -36,14 +36,51 @@ Multi-modality (MM) image fusion aims to render fused images that maintain the m
 
 Our CDDFuse is implemented in ``net.py``.
 
+### Training
+**1. Virtual Environment**
+```
+# create virtual environment
+conda create -n cddfuse python=3.8.10
+conda activate cddfuse
+# select pytorch version yourself
+# install cddfuse requirements
+pip install -r requirements.txt
+```
+
+**2. Data Preparation**
+
+Download the MSRS dataset from [this link](https://github.com/Linfeng-Tang/MSRS) and place it in the folder ``'./MSRS_train/'``.
+
+**3. Pre-Processing**
+
+Run 
+```
+python dataprocessing.py
+``` 
+and the processed training dataset is in ``'./data/MSRS_train_imgsize_128_stride_200.h5'``.
+
+**4. CDDFuse Training**
+
+Run 
+```
+python train.py
+``` 
+and the trained model is available in ``'./models/'``.
+
 ### Testing
 
+**1. Pretrained models**
+
 Pretrained models are available in ``'./models/CDDFuse_IVF.pth'`` and ``'./models/CDDFuse_MIF.pth'``, which are responsible for the Infrared-Visible Fusion (IVF) and Medical Image Fusion (MIF) tasks, respectively. 
 
+**2. Test datasets**
+
 The test datasets used in the paper have been stored in ``'./test_img/RoadScene'``, ``'./test_img/TNO'`` for IVF, ``'./test_img/MRI_CT'``, ``'./test_img/MRI_PET'`` and ``'./test_img/MRI_SPECT'`` for MIF.
 
 Unfortunately, since the size of **MSRS dataset** for IVF is 500+MB, we can not upload it for exhibition. It can be downloaded via [this link](https://github.com/Linfeng-Tang/MSRS). The other datasets contain all the test images.
 
+**3. Results in Our Paper**
+
 If you want to infer with our CDDFuse and obtain the fusion results in our paper, please run ``'test_IVF.py'`` for IVF and ``'test_MIF.py'`` for MIF. 
 
 The testing results will be printed in the terminal. 
@@ -91,38 +128,6 @@ CDDFuse_MIF     3.9     58.31   20.87   2.49    1.35    0.97    0.78    1.48
 ```
 which can match the results in Table 5 in our original paper.
 
-### Training
-**1. Virtual Environment**
-```
-# create virtual environment
-conda create -n cddfuse python=3.8.10
-conda activate cddfuse
-# select pytorch version yourself
-# install cddfuse requirements
-pip install -r requirements.txt
-```
-
-**2. Data Preparation**
-
-Download the MSRS dataset from [this link](https://github.com/Linfeng-Tang/MSRS) and place it in the folder ``'./MSRS_train/'``.
-
-**3. Pre-Processing**
-
-Run 
-```
-python dataprocessing.py
-``` 
-and the processed training dataset is in ``'./data/MSRS_train_imgsize_128_stride_200.h5'``.
-
-**4. CDDFuse Training**
-
-Run 
-```
-python train.py
-``` 
-and the trained model is available in ``'./models/'``.
-
-
 ## CDDFuse
 
 ### Illustration of our CDDFuse model.