deep_complex_networks-master
所属分类：数据挖掘/数据仓库
开发工具：Python
文件大小：160KB
下载次数：3
上传日期：2017-10-20 10:21:51
上 传 者：
木子日辰
说明： 该存储库包含重现深层复杂网络文章中提供的实验的代码。
(This repository contains code which reproduces experiments presented in the paper Deep Complex Networks.)
文件列表:
LICENSE.md (1367, 2017-09-01)
complexnn (0, 2017-09-01)
complexnn\__init__.py (861, 2017-09-01)
complexnn\bn.py (21518, 2017-09-01)
complexnn\conv.py (43085, 2017-09-01)
complexnn\dense.py (9107, 2017-09-01)
complexnn\fft.py (5578, 2017-09-01)
complexnn\init.py (10779, 2017-09-01)
complexnn\norm.py (11241, 2017-09-01)
complexnn\pool.py (4876, 2017-09-01)
complexnn\utils.py (2186, 2017-09-01)
imgs (0, 2017-09-01)
imgs\precision_recall.png (37138, 2017-09-01)
imgs\pred_gt.png (27283, 2017-09-01)
musicnet (0, 2017-09-01)
musicnet\musicnet (0, 2017-09-01)
musicnet\musicnet\__init__.py (0, 2017-09-01)
musicnet\musicnet\callbacks.py (3639, 2017-09-01)
musicnet\musicnet\dataset.py (7392, 2017-09-01)
musicnet\musicnet\models (0, 2017-09-01)
musicnet\musicnet\models\__init__.py (4162, 2017-09-01)
musicnet\musicnet\models\complex (0, 2017-09-01)
musicnet\musicnet\models\complex\__init__.py (3946, 2017-09-01)
musicnet\scripts (0, 2017-09-01)
musicnet\scripts\resample.py (1751, 2017-09-01)
musicnet\scripts\train.py (4449, 2017-09-01)
notebooks (0, 2017-09-01)
notebooks\visualize_musicnet.ipynb (90053, 2017-09-01)
scripts (0, 2017-09-01)
scripts\run.py (9102, 2017-09-01)
scripts\run.sh (155, 2017-09-01)
scripts\training.py (24414, 2017-09-01)
setup.py (576, 2017-09-01)
Deep Complex Networks
=====================
This repository contains code which reproduces experiments presented in
the paper [Deep Complex Networks](https://arxiv.org/abs/1705.09792).
Requirements
------------
Install requirements for computer vision experiments with pip:
```
pip install numpy Theano keras kerosene
```
And for music experiments:
```
pip install scipy sklearn intervaltree resampy
pip install git+git://github.com/bartvm/mimir.git
```
Depending on your Python installation you might want to use anaconda or other tools.
Installation
------------
```
python setup.py install
```
Experiments
-----------
### Computer vision
1. Get help:
```
python scripts/run.py train --help
```
2. Run models:
```
python scripts/run.py train -w WORKDIR --model {real,complex} --sf STARTFILTER --nb NUMBEROFBLOCKSPERSTAGE
```
Other arguments may be added as well; Refer to run.py train --help for
- Optimizer settings
- Dropout rate
- Clipping
- ...
### MusicNet
0. Download the dataset from [the official page](https://homes.cs.washington.edu/~thickstn/musicnet.html)
```
mkdir data/
wget https://homes.cs.washington.edu/~thickstn/media/musicnet.npz -P data/
```
1. Resample the dataset with
```
resample.py data/musicnet.npz data/musicnet_11khz.npz 44100 11000
```
2. Run shallow models
```
train.py shallow_model --in-memory --model=shallow_convnet --local-data data/musicnet_11khz.npz
train.py shallow__complex_model --in-memory --model=complex_shallow_convnet --complex --local-data data/musicnet_11khz.npz
```
3. Run deep models
```
train.py deep_model --in-memory --model=deep_convnet --fourier --local-data data/musicnet_11khz.npz
train.py deep_complex_model --in-memory --model=complex_deep_convnet --fourier --complex --local-data data/musicnet_11khz.npz
```
4. Visualize with jupyter notebook
Run the notebook `notebooks/visualize_musicnet.ipynb`.
Citation
--------
Please cite our work as
```
@ARTICLE {,
author = "Chiheb Trabelsi, Olexa Bilaniuk, Dmitriy Serdyuk, Sandeep Subramanian, Joo Felipe Santos, Soroush Mehri, Negar Rostamzadeh, Yoshua Bengio, Christopher J Pal",
title = "Deep Complex Networks",
journal = "arXiv preprint arXiv:1705.09792",
year = "2017"
}
```
近期下载者：
相关文件：
收藏者：