reorganized to make the important stuff easier to find

index.html

@@ -134,24 +134,7 @@ <h1>An Interactive Node-Link Visualization of Convolutional Neural Networks</h1>
 	  </div>
 	  <h2>Abstract</h2>
 	  <p>Convolutional neural networks are at the core of state-of-the-art approaches to a variety of computer vision tasks. Visualizations of neural networks typically take the form of static node-link diagrams, which illustrate only the structure of a network, rather than the behavior. Motivated by this observation, this paper presents a new interactive visualization of neural networks trained on handwritten digit recognition, with the intent of showing the actual behavior of the network given user-provided input. The user can interact with the network through a drawing pad, and watch the activation patterns of the network respond in real time.</p>
-	  <h2>Paper</h2>
-	  <div id="pdf">
-	    <a href="harley_vis_isvc15.pdf"><img src="images/paper_noshadow.png" alt="An Interactive Node-Link Visualization of Convolutional Neural Networks"></a>
-	  </div>
-	  <h2>Citation</h2>
-	  <div class="cite">A. W. Harley, "An Interactive Node-Link Visualization of Convolutional Neural Networks," in ISVC, pages 867-877, 2015</div>
-	  <p>Bibtex format:</p>
-	  <div class="cite">
-	    <code>
-	      @inproceedings{harley2015isvc,<br/>
-	      &nbsp;&nbsp;&nbsp;&nbsp;title = {An Interactive Node-Link Visualization of Convolutional Neural Networks},<br/>
-	      &nbsp;&nbsp;&nbsp;&nbsp;author = {Adam W Harley},<br/>
-	      &nbsp;&nbsp;&nbsp;&nbsp;booktitle = {ISVC},<br/>
-	      &nbsp;&nbsp;&nbsp;&nbsp;pages = {867--877},<br/>
-	      &nbsp;&nbsp;&nbsp;&nbsp;year = {2015}<br/>
-	      }
-	    </code>
-	  </div>
+
 	  <h2>Demo</h2>
 	  <div class="projects">
 	    <div class="project">				
@@ -199,8 +182,34 @@ <h2>Demo</h2>
 	      </table>
 	    </div>
 	  </div>			
-	  <h2>Details</h2>
-	  <p>The networks were trained on an augmented version of MNIST, so they excel at categorizing centred upright numbers. The networks were trained in a custom neural network implementation in MATLAB; the math for the visualizations was written in Javascript; the visualization was created in WebGL. The source code for both visualizations is available <a href="https://github.com/aharley/nn_vis">here</a>.</p>
+	  <h2>Code and implementation details</h2>
+	  <p>Our source code is freely available on github <a href="https://github.com/aharley/nn_vis">here</a>.</p>
+	  <p>A few details that might be interesting:
+	    <ul>
+	      <li>The neural networks were implemented and trained using a custom MATLAB implementation of neural networks and backpropagation, which I wrote to help myself learn the basics.</li>
+	      <li>The neural networks' training data was an augmented version of MNIST, so as a result, these networks do best when the input drawings are upright and centered.</li>
+	      <li>The math for the visualizations is written in Javascript.</li>
+	      <li>The visualization itself is written in WebGL.</li>
+	    </ul>	      
+	  <h2>Paper</h2>
+	  <div id="pdf">
+	    <a href="harley_vis_isvc15.pdf"><img src="images/paper_noshadow.png" alt="An Interactive Node-Link Visualization of Convolutional Neural Networks"></a>
+	  </div>
+	  <h2>Citation</h2>
+	  <div class="cite">A. W. Harley, "An Interactive Node-Link Visualization of Convolutional Neural Networks," in ISVC, pages 867-877, 2015</div>
+	  <p>Bibtex format:</p>
+	  <div class="cite">
+	    <code>
+	      @inproceedings{harley2015isvc,<br/>
+	      &nbsp;&nbsp;&nbsp;&nbsp;title = {An Interactive Node-Link Visualization of Convolutional Neural Networks},<br/>
+	      &nbsp;&nbsp;&nbsp;&nbsp;author = {Adam W Harley},<br/>
+	      &nbsp;&nbsp;&nbsp;&nbsp;booktitle = {ISVC},<br/>
+	      &nbsp;&nbsp;&nbsp;&nbsp;pages = {867--877},<br/>
+	      &nbsp;&nbsp;&nbsp;&nbsp;year = {2015}<br/>
+	      }
+	    </code>
+	  </div>
+
 	</div>
       </div>
     </div>