Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/tensorpro/FractalNet
Implementation of FractalNet.
https://github.com/tensorpro/FractalNet
Last synced: 8 days ago
JSON representation
Implementation of FractalNet.
- Host: GitHub
- URL: https://github.com/tensorpro/FractalNet
- Owner: tensorpro
- License: mit
- Created: 2016-06-29T23:39:52.000Z (over 8 years ago)
- Default Branch: master
- Last Pushed: 2017-01-26T02:18:24.000Z (almost 8 years ago)
- Last Synced: 2024-08-01T22:49:55.225Z (3 months ago)
- Language: Python
- Homepage: https://arxiv.org/pdf/1605.07648v1.pdf
- Size: 737 KB
- Stars: 48
- Watchers: 4
- Forks: 12
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
- awesome-image-classification - unofficial-tensorflow : https://github.com/tensorpro/FractalNet
- awesome-image-classification - unofficial-tensorflow : https://github.com/tensorpro/FractalNet
README
# FractalNet
Implementation of FractalNet (https://arxiv.org/pdf/1605.07648v1.pdf) in TensorFlow.Fractal blocks have components with separated columns, as well as the fractal layout.
For rounds of training with global drop path, the separated layout is used.
In all other cases, the fractal layout is used.#Fractal
![Fractal Visualization](imgs/Fractal.png)#Separated
![Separated Visualization](imgs/Seperated.png)#Entire Block
![Block Visualization](imgs/FractalBlock.png)