[Example] Gradcam consolidation in tutorial

docs/conf.py

@@ -107,7 +107,7 @@
 
 # List of patterns, relative to source directory, that match files and
 # directories to ignore when looking for source files.
-exclude_patterns = ['3rdparty', 'build_version_doc', 'virtualenv', 'api/python/model.md', 'README.md']
+exclude_patterns = ['3rdparty', 'build_version_doc', 'virtualenv', 'api/python/model.md', 'README.md', 'tutorial_utils']
 
 # The reST default role (used for this markup: `text`) to use for all documents.
 #default_role = None

docs/tutorials/vision/cnn_visualization.md

@@ -22,7 +22,7 @@ from matplotlib import pyplot as plt
 import numpy as np
 
 gradcam_file = "gradcam.py" 
-base_url = "https://raw.githubusercontent.com/indhub/mxnet/cnnviz/example/cnn_visualization/{}?raw=true"
+base_url = "https://github.com/apache/incubator-mxnet/tree/master/docs/tutorial_utils/vision/cnn_visualization/{}?raw=true"
 mx.test_utils.download(base_url.format(gradcam_file), fname=gradcam_file)
 import gradcam
 ```
@@ -182,6 +182,7 @@ Next, we'll write a method to get an image, preprocess it, predict category and
 2. **Guided Grad-CAM:** Guided Grad-CAM shows which exact pixels contributed the most to the CNN's decision.
 3. **Saliency map:** Saliency map is a monochrome image showing which pixels contributed the most to the CNN's decision. Sometimes, it is easier to see the areas in the image that most influence the output in a monochrome image than in a color image.
 
+
 ```python
 def visualize(net, img_path, conv_layer_name):
     orig_img = mx.img.imread(img_path)