diff --git a/README.md b/README.md
index 565ff1b..b805ad7 100644
--- a/README.md
+++ b/README.md
@@ -1,9 +1,7 @@
-Crossnet, a cross-view supervised learning solution
+Predicting Ground-Level Scene Layout from Aerial Imagery
 ========
 
-This code is a tensorflow implementation of this paper, [Predicting Ground-Level Scene Layout from Aerial Imagery](http://openaccess.thecvf.com/content_cvpr_2017/papers/Zhai_Predicting_Ground-Level_Scene_CVPR_2017_paper.pdf).
-
-And you are welcome to visit our [Project Page](http://cs.uky.edu/~ted/research/crossview/)
+This code is a TensorFlow implementation of this CVPR 2017 paper: [Predicting Ground-Level Scene Layout from Aerial Imagery](http://openaccess.thecvf.com/content_cvpr_2017/papers/Zhai_Predicting_Ground-Level_Scene_CVPR_2017_paper.pdf).
 
 Dependencies
 ------------
@@ -22,9 +20,5 @@ $ python main.py --is_training=False
 ```
 Data
 ------------
-This repo only contains some example data for training, the whole dataset can be found in [this link](https://drive.google.com/open?id=0BzvmHzyo_zCAX3I4VG1mWnhmcGc).
-You would have to edit the dataset path in main.py in order to use it.
+This repository only contains example data for training; the full dataset can be requested from [this link]([https://drive.google.com/open?id=0BzvmHzyo_zCAX3I4VG1mWnhmcGc](https://mvrl.cse.wustl.edu/datasets/cvusa/). You would have to edit the dataset path in main.py in order to use it.
 
-Note
-------------
-We plan to release evaluation code soon.