make network loading instructions clearer

Author: Clément Pinard

README.md

@@ -17,8 +17,12 @@ There is currently no implementation of FlowNetC as a specific Correlation layer
 
 ## Pretrained Models
 Thanks to [Kaixhin](https://github.com/Kaixhin) you can download a pretrained version of FlowNetS (from caffe, not from pytorch) [here](https://drive.google.com/open?id=0B5EC7HMbyk3CbjFPb0RuODI3NmM). This folder also contains trained networks from scratch.
+
+### Note on networks loading
+Directly feed the downloaded Network to the script, you don't need to uncompress it even if your desktop environment tells you so.
+
 ### Note on networks from caffe
-These networks expect a BGR input in range [-0.5,0.5]. However, BGR order is not very important.
+These networks expect a BGR input in range [-0.5,0.5] (compared to RGB in pytorch). However, BGR order is not very important.
 
 ## Prerequisite