zcdliuwei / CenterNet_TensorFlow2

A tensorflow2.x implementation of CenterNet.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

CenterNet_TensorFlow2

A tensorflow2.x implementation of CenterNet.

Requirements:

  • Python >= 3.6
  • TensorFlow == 2.2.0rc3
  • numpy
  • opencv-python

Results

The following are the detection results of some pictures in the PASCAL VOC 2012 dataset.
img_1
img_2
img_3

Usage

Train on PASCAL VOC 2012

  1. Download the PASCAL VOC 2012 dataset.
  2. Unzip the file and place it in the 'data/datasets' folder, make sure the directory is like this :
|——data
    |——datasets
        |——VOCdevkit
            |——VOC2012
                |——Annotations
                |——ImageSets
                |——JPEGImages
                |——SegmentationClass
                |——SegmentationObject
  1. Run write_to_txt.py to generate data.txt.
  2. Run train.py to start training, before that, you can change the value of the parameters in configuration.py.

Test on single picture

  1. Change the test_single_image_dir in configuration.py.
  2. Run test.py to test on single picture.

Acknowledgments

  1. Official PyTorch implementation of CenterNet: https://github.com/xingyizhou/CenterNet
  2. A TensorFlow implementation of CenterNet: https://github.com/MioChiu/TF_CenterNet

References

  1. Objects as Points
ezoic increase your site revenue

About

A tensorflow2.x implementation of CenterNet.

License:MIT License


Languages

Language:Python 100.0%