nn_vis - GitLite

KO | EN

gitlite — search

#python #hacktoberfest #react #ai #golang #machine-learning #rust #deep-learning #awesome #kubernetes #vue #pytorch #openai #open-source

nn_vis

(★ 444)

An interactive visualization of neural networks

파일 탐색기

최종 버전 다운로드 (.zip)

- 2d.html
- 3d.html
- drawingboard2.css
- katex.min.css
- main.css
- clear64.png
- cnn_2d.png
- cnn_3d.png
- convnet_480.png
- convnet_flat_480.png
- eraser64.png
- fc_flat_480.png
- fcnet_480.png
- mlp_2d.png
- mlp_3d.png
- paper.png
- paper_noshadow.png
- pencil64.png
- popsci_graphic_flat.png
- seven.png
- - myColorMap_dark.js
  - myColorMap_gray.js
  - myColorMap_light.js
- - board.js
  - utils.js
- - droid_sans_regular.typeface.js
- - 300_100_tanh_aug_2pass.js
  - common.js
  - conv.js
  - convnet_weights.js
  - fc.js
  - webgl_784_300_100_positions.json
  - webgl_784_300_100_positions_flat.json
  - webgl_convnet2.json
  - webgl_convnet_flat.json
- camera.js
- createText.js
- droid_sans_regular.typeface.js
- GeometryUtils.js
- jquery-1.11.2.min.js
- katex.min.js
- math.min.js
- myOrbitControls.js
- processing.min.js
- stats.min.js
- sylvester.src.js
- three.min.js
- tween.min.js
- 2d.html
- 3d.html
.gitignore
harley_vis_isvc15.pdf
index.html
LICENSE
README.md

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# An Interactive Node-Link Visualization of Convolutional Neural Networks
![alt text](https://raw.githubusercontent.com/aharley/nn_vis/master/images/seven.png)
## Abstract
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 project 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.

## Paper PDF

<a href="https://adamharley.com/nn_vis/harley_vis_isvc15.pdf" rel="paper">![paper](https://raw.githubusercontent.com/aharley/nn_vis/master/images/paper.png)</a>

## Live demos
Live demos for all models are available at [https://adamharley.com/nn_vis](https://adamharley.com/nn_vis):

1. 3d visualization of a multi-layer perceptron:

<a href="https://adamharley.com/nn_vis/mlp/3d.html" rel="mlp_3d">![cnn2d](https://raw.githubusercontent.com/aharley/nn_vis/master/images/mlp_3d.png)</a>

2. 3d visualization of a convolutional network:

<a href="https://adamharley.com/nn_vis/cnn/3d.html" rel="cnn_3d">![cnn2d](https://raw.githubusercontent.com/aharley/nn_vis/master/images/cnn_3d.png)</a>

3. 2d visualization of a multi-layer perceptron:

<a href="https://adamharley.com/nn_vis/mlp/2d.html" rel="mlp_2d">![cnn2d](https://raw.githubusercontent.com/aharley/nn_vis/master/images/mlp_2d.png)</a>

4. 2d visualization of a convolutional network:

<a href="https://adamharley.com/nn_vis/cnn/2d.html" rel="cnn_2d">![cnn2d](https://raw.githubusercontent.com/aharley/nn_vis/master/images/cnn_2d.png)</a>

## FAQ

* Can I use this in my course/textbook/presentation?
  * Yes! Please just cite the work appropriately.
  
* How do I cite you?
  * Here is a plaintext citation:
  
    `A. W. Harley, "An Interactive Node-Link Visualization of Convolutional Neural Networks," in ISVC, pages 867-877, 2015`
  
    Here is a bibtex snippet:
    ```
    @inproceedings{harley2015isvc,
    title = {An Interactive Node-Link Visualization of Convolutional Neural Networks},
    author = {Adam W Harley},
    booktitle = {ISVC},
    pages = {867--877},
    year = {2015}
    }
    ```
    
* I tried to run the code locally, and I see classifications, but I do not see the network visualization. What's wrong?
  * This is probably related to json requests being blocked by something called `CORS policy`. The solution is to upload the code to a web address and run it from there, instead of running locally.

Contact: aharley@cs.stanford.edu