https://github.com/huybery/visualizingcnn
:see_no_evil:A PyTorch implementation of the paper "Visualizing and Understanding Convolutional Networks." （ECCV 2014)
https://github.com/huybery/visualizingcnn
computer-vision convolutional-neural-networks deeplearning
Last synced: 10 months ago
JSON representation
:see_no_evil:A PyTorch implementation of the paper "Visualizing and Understanding Convolutional Networks." （ECCV 2014)
- Host: GitHub
- URL: https://github.com/huybery/visualizingcnn
- Owner: huybery
- License: mit
- Created: 2018-05-08T05:36:15.000Z (about 8 years ago)
- Default Branch: master
- Last Pushed: 2018-08-05T08:33:00.000Z (almost 8 years ago)
- Last Synced: 2024-12-09T23:51:19.478Z (over 1 year ago)
- Topics: computer-vision, convolutional-neural-networks, deeplearning
- Language: Python
- Homepage: https://arxiv.org/abs/1311.2901
- Size: 414 KB
- Stars: 197
- Watchers: 4
- Forks: 38
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# VisualizingCNN
A PyTorch implementation of the 2014 ECCV paper "Visualizing and understanding convolutional networks"
```bash
Predicted: [('n02123045', 'tabby', 0.5042504668235779), ('n02124075', 'Egyptian_cat', 0.26163962483406067), ('n02123159', 'tiger_cat', 0.23190157115459442)]
```
## Usage
```bash
python main.py
```
## Requirement
```bash
Pytorch == 0.4.0
opencv-python == 3.4.0.12
```
## Detail
In original paper, author shows the top 9 activations in a random subset of eature maps across the validation data, projected down to pixel space using there deconvolutional network approach. But in this project, we only show the max activations (top 1) for each layer projected down to pixel space by the single image.
## Notes
The network use vgg16 pretrained from torchvision.models, the reconstruction proposal is human's labeling, rather model generate.