diff --git a/.DS_Store b/.DS_Store
index 66d3f60..56d349c 100644
Binary files a/.DS_Store and b/.DS_Store differ
diff --git a/README.md b/README.md
index 2611ba1..b370277 100644
--- a/README.md
+++ b/README.md
@@ -1,12 +1,22 @@
# PSVH-3d-reconstruction
-This repository is the implementation of our AAAI19 paper [Deep Single-View 3D Object Reconstruction with Visual Hull Embedding](https://arxiv.org/pdf/1809.03451.pdf).
+This repository is the implementation of our AAAI 2019 paper:
-The authors of this paper are [Hanqing Wang](https://qweas120.github.io), [Jiaolong Yang](http://jlyang.org/), [Wei Liang](http://iitlab.bit.edu.cn/mcislab/~liangwei/) and [Xin Tong](http://www.xtong.info/).
+[Deep Single-View 3D Object Reconstruction with Visual Hull Embedding](https://arxiv.org/pdf/1809.03451.pdf)
+[Hanqing Wang](https://qweas120.github.io), [Jiaolong Yang](http://jlyang.org/), [Wei Liang](http://iitlab.bit.edu.cn/mcislab/~liangwei/), [Xin Tong](http://www.xtong.info/)
+
+This work is implemented using [TensorFlow](https://www.tensorflow.org/).
## Introduction
+In this paper, we present an approach which aims to preserve more shape details and improve the reconstruction quality. The key idea of our method is to lever- age object mask and pose estimation from CNNs to assist the 3D shape learning by constructing a probabilistic single-view visual hull inside of the network.
+
+
+Our method works by first predicting a coarse shape as well as the object pose and silhouette using CNNs, followed by a novel 3D refinement CNN which refines the coarse shapes using the constructed probabilistic visual hulls.
-
+
+
+## Examples
+
## Citation
If you find our work helpful for your research, please cite our paper:
@@ -31,7 +41,7 @@ conda install tensorflow pillow
```
The checkpoint of the trained models are available [here](https://drive.google.com/open?id=1TJEUUhmZL8WJgQbsrRX9D_GKAiqE8Gic)(426MB). Extract the files to the root directory.
-## Run Example
+## Demo
Run `python run_case.py` to run the examples. The outputs are reconstruction results before and after the refinement (Please refer to our paper for more details). The results are in `obj` format. You can use [meshlab](http://www.meshlab.net/) for visulization.
@@ -39,3 +49,5 @@ Run `python run_case.py` to run the examples. The outputs are reconstruction res
+## License
+PSVH is freely available for non-commercial use, and may be redistributed under these conditions. Please see the license for further details. For commercial license, please contact the authors.
\ No newline at end of file
diff --git a/readme/.DS_Store b/readme/.DS_Store
index 5bf9d70..90d14b1 100644
Binary files a/readme/.DS_Store and b/readme/.DS_Store differ
diff --git a/readme/img/.DS_Store b/readme/img/.DS_Store
index 8e27d80..dc341a1 100644
Binary files a/readme/img/.DS_Store and b/readme/img/.DS_Store differ
diff --git a/readme/img/network.png b/readme/img/network.png
new file mode 100644
index 0000000..1138f28
Binary files /dev/null and b/readme/img/network.png differ
diff --git a/readme/img/results/result.png b/readme/img/results/result.png
index 91a33f3..c579aaf 100644
Binary files a/readme/img/results/result.png and b/readme/img/results/result.png differ