https://github.com/tusimple/sparse-structure-selection
https://github.com/tusimple/sparse-structure-selection
Last synced: 10 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/tusimple/sparse-structure-selection
- Owner: TuSimple
- License: apache-2.0
- Created: 2018-04-19T05:11:01.000Z (almost 8 years ago)
- Default Branch: master
- Last Pushed: 2018-07-17T03:47:04.000Z (over 7 years ago)
- Last Synced: 2025-04-17T03:05:14.428Z (10 months ago)
- Language: Python
- Size: 277 KB
- Stars: 87
- Watchers: 8
- Forks: 21
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# sparse-structure-selection
This code is a re-implementation of the imagenet classification experiments in the paper [Data-Driven Sparse Structure Selection for Deep Neural Networks
](https://arxiv.org/abs/1707.01213) (ECCV2018).
## Citation
If you use our code in your research or wish to refer to the baseline results, please use the following BibTeX entry.
```
@article{SSS2018
author = {Zehao Huang and Naiyan Wang},
title = {Data-Driven Sparse Structure Selection for Deep Neural Networks},
journal = {ECCV},
year = {2018}
}
```
## Implementation
This code is implemented by a modified [MXNet](https://github.com/huangzehao/incubator-mxnet-bk) which supports [ResNeXt-like](https://github.com/facebookresearch/ResNeXt) augmentation. (This version of MXNet does not support cudnn7)
## ImageNet data preparation
Download the [ImageNet](http://image-net.org/download-images) dataset and create pass through rec (following [tornadomeet's repository](https://github.com/tornadomeet/ResNet#imagenet) but using unchange mode)
## Run
- modify ```config/cfgs.py```
- ```python train.py```
## Results on ImageNet-1k
