https://github.com/uranusx86/random-erasing-tensorflow

A complete Tensorflow implementation of cutout random erasing (without numpy)
https://github.com/uranusx86/random-erasing-tensorflow

argumentation deep-learning image-processing neural-network paper-implementations preprocessing python tensorflow tensorflow-experiments

Last synced: 3 months ago
JSON representation

A complete Tensorflow implementation of cutout random erasing (without numpy)

• Host: GitHub
• URL: https://github.com/uranusx86/random-erasing-tensorflow
• Owner: uranusx86
• Created: 2018-12-14T13:02:58.000Z (about 6 years ago)
• Default Branch: master
• Last Pushed: 2019-01-08T12:17:04.000Z (about 6 years ago)
• Last Synced: 2024-09-22T08:02:06.261Z (4 months ago)
• Topics: argumentation, deep-learning, image-processing, neural-network, paper-implementations, preprocessing, python, tensorflow, tensorflow-experiments
• Language: Python
• Homepage:
• Size: 529 KB
• Stars: 13
• Watchers: 2
• Forks: 6
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Random-Erasing-tensorflow

A data argumentation implementation of random erasing using Tensorflow 


WITHOUT numpy !

**Note: this side project is just for fun, the performance is equal to numpy version when using i5 CPU, but a bit lower than numpy version when using i7 CPU**

# Dependency

1. Tensorflow 1.10+

# Result

![](https://github.com/uranusx86/Random-Erasing-tensorflow/blob/master/data/random_erasing.jpg)

# Test Performance

```

python3 test_performance.py

```

Current version test @ CPU i5-3470 & GPU GTX1080ti 


numpy: 0.1695 sec/image 


TF: 0.1696 sec/image

# Ref

1. [Random Erasing Data Augmentation](http://arxiv.org/abs/1602.02830), Zhun Zhong, Liang Zheng, Guoliang Kang, Shaozi Li, Yi Yang

2. [Random Erasing (Author version)](https://github.com/zhunzhong07/Random-Erasing)

3. [Cutout Random Erasing (Keras version)](https://github.com/yu4u/cutout-random-erasing)