Update README.md

README.md

@@ -1,6 +1,6 @@
 # To Run the Network on Test Data:
 
--Step 1: Running the Network to Detect the ON and OFF surfaces 
+## -Step 1: Running the Network to Detect the ON and OFF surfaces 
 + First, put all the images to be processed (STD Tiff stacks of chAT Bands) into the ImagesHere folder.
 
 + Then, open Matlab, make sure Matlab has included all the folders and subfolders inside of VNet, and run the script RunMe.m
@@ -11,7 +11,7 @@
 
 ######################################################################################################
 
--Step 2: Clean Up
+## -Step 2: Clean Up
 + Make sure to delete or remove the images in ImageHere so that you can use the network again with other images.
 
 + Delete all the Detected Surfaces in DetectedSurfaces that are no longer needed also.
@@ -24,12 +24,12 @@ Notes: In the case where the overlays are not correct, manually annotated the da
 
 # To Train the Network with new Data:
 
--Step 1: Install Dependencies and Libraries 
+## -Step 1: Install Dependencies and Libraries 
 + Run CaffeInstallation.sh and PythonLibraryInstallation.sh. To run the 2 files, either double-click on the file name and choose 'Run On Terminal', or manually open up a new Terminal, change directory (cd command) to locate the file, and type:
             ./filename.sh 
 in order to execute the command that download and install dependencies necessary to run pycaffe.
 
--Step 2: Prepare training data
+## -Step 2: Prepare training data
 + The ON and OFF band should be trained on separated network for best result, and all the images has to be resized to 128x128x64 before feeding into the network.
 
 + First, put all the raw images that needed to be train in the Dataset/RawImages folder.
@@ -38,7 +38,7 @@ in order to execute the command that download and install dependencies necessary
 
 + Find the CreateTrainingData.m script and run it on Matlab to create the Groundtruths, resize the images and the groundtruths, and put them in the right location ready to be trained.
 
--Step 3: Run the training script
+## -Step 3: Run the training script
 + Open a Terminal, locate the VNet Directory, and type: python main.py -trainON to run and train the network with the ON chATband data. This should take 2-3 days depending on the number of iterations specified in main.py (default is 100,000).
 
 + After the network has completed training the ON data, open a new Terminal, locate the VNet Directory, and type: python main.py -trainOFF to run and train the network with the OFF chATband data. This should also take 2-3 days depending on the number of iterations specified in main.py (default is 100,000).