https://github.com/mingtaoguo/retinanet_tensorflow

Implementation of RetinaNet (focal loss) by TensorFlow (object detection)
https://github.com/mingtaoguo/retinanet_tensorflow

focal-loss object-detection retinanet tensorflow

Last synced: 12 months ago
JSON representation

Implementation of RetinaNet (focal loss) by TensorFlow (object detection)

• Host: GitHub
• URL: https://github.com/mingtaoguo/retinanet_tensorflow
• Owner: MingtaoGuo
• License: mit
• Created: 2019-11-29T02:08:07.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2019-11-29T03:22:24.000Z (over 6 years ago)
• Last Synced: 2025-04-05T13:11:21.118Z (about 1 year ago)
• Topics: focal-loss, object-detection, retinanet, tensorflow
• Language: Python
• Homepage:
• Size: 664 KB
• Stars: 16
• Watchers: 1
• Forks: 2
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# RetinaNet_TensorFlow

Implementation of RetinaNet by TensorFlow (object detection)

# Introduction

![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/introduction.jpg)

### Focal Loss

![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/formula.jpg)

``` python

def focal_loss(logits, labels, alpha=0.25, gamma=2):

    pos_pt = tf.clip_by_value(tf.nn.sigmoid(logits), 1e-10, 0.999)

    fl = labels * tf.log(pos_pt) * tf.pow(1 - pos_pt, gamma) * alpha + (1 - labels) * tf.log(1 - pos_pt) * tf.pow(pos_pt, gamma) * (1 - alpha)

    fl = -tf.reduce_sum(fl, axis=2)

    return fl

```

# How to use

### Dataset

Pascal Voc: http://pjreddie.com/media/files/VOCtrainval_06-Nov-2007.tar

### Training phase

1. Downloading the pre-trained model of ResNet50, and put it into the folder **resnet_ckpt** 

   

   Address: http://download.tensorflow.org/models/resnet_v2_50_2017_04_14.tar.gz

2. According to your own dataset, modify the **config.py** by yourself

3. Executing **train.py** 

### Testing phase

1. Changing the IMG_PATH or VIDEO_PATH in **test.py** for testing

2. Executing **test.py**

    Model we trained: https://drive.google.com/open?id=1_j-bjQ_SWT3txqCabM8_Ny0IMUqTb8Lk

# Requirement

1. python

2. tensorflow

3. pillow

4. numpy

5. cv2

# Results

|Total Loss|Class Loss|Box Loss|

|-|-|-|

|![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/total_loss.jpg)|![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/class_loss.jpg)|![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/box_loss.jpg)|

||||

|-|-|-|

|![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/1.jpg)|![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/8.jpg)|![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/6.jpg)|

|![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/7.jpg)|![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/9.jpg)|![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/4.jpg)|

|![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/2.jpg)|![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/3.jpg)|![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/14.jpg)|

|![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/10.jpg)|![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/12.jpg)|![](https://github.com/MingtaoGuo/RetinaNet_TensorFlow/blob/master/IMGS/13.jpg)|

# Reference

[1] Lin, Tsung-Yi, et al. "Focal loss for dense object detection." Proceedings of the IEEE international conference on computer vision. 2017.

# Author

Mingtao Guo