New installation and minor fixes for theano 1.0.2

Author: chrischoy

.theanorc

@@ -1,3 +1,3 @@
 [global]
 floatX = float32
-device = gpu0
+device = cuda

README.md

@@ -60,17 +60,22 @@ The package requires python3. You can follow the direction below to install virt
 git clone https://github.com/chrischoy/3D-R2N2.git
 ```
 
-- Setup virtual environment and install requirements and copy the theanorc file to the `$HOME` directory
+- Setup the anaconda virtual environment and installing requirements ([How to use anaconda](https://conda.io/docs/user-guide/install/index.html))
 
 ```
 cd 3D-R2N2
-pip install virtualenv
-virtualenv -p python3 --system-site-packages py3
-source py3/bin/activate
+conda create -n py3-theano python=3.6
+source activate py3-theano
+conda install pygpu
 pip install -r requirements.txt
+```
+- copy the theanorc file to the `$HOME` directory
+
+```
 cp .theanorc ~/.theanorc
 ```
 
+
 ### Running demo.py
 
 - Install meshlab (skip if you have another mesh viewer). If you skip this step, demo code will not visualize the final prediction.

lib/layers.py

@@ -319,12 +319,11 @@ def set_output(self):
             padded_input = self._prev_layer.output
             padded_input_shape = self._input_shape
 
-        conv_out = conv.conv2d(
+        conv_out = tensor.nnet.conv2d(
             input=padded_input,
             filters=self.W.val,
             filter_shape=self._filter_shape,
-            image_shape=np.asarray(
-                padded_input_shape, dtype=np.int16),
+            input_shape=padded_input_shape,
             border_mode='valid')
 
         # add the bias term. Since the bias is a vector (1D array), we first