Keras Object Detection API

You Only Look Keras

As a 2019 Open Source Countibuthon Project, we create Kears Object Detection API.
The purpose of Keras Object Detection API is to make it easy and fast so that everyone can create their own Object Detection model without difficulty. Detailed instructions for use will be updated at a later. You can look forward to it. 🤓

Directory Structure
Installation
Tutorial
Quick Start
Dependencies
Release information
Contributors

Directory Structure

KerasObjectDetector
├── README.md
├── setup.py
├── setup.md
├── docker files
├── datasets
├── utils
│   ├── image1.png
│   ├── image2.jpeg
│   ├── result_image.png
│   ├── show_bbox.py
│   ├── test.py...

Installation (On Linux)

First, Download YOLK API that help to set up development environment for working on object detection. Enter the following command in terminal.

  # Download YOLK API
  $ git clone https://github.com/KerasKorea/KerasObjectDetector.git
  $ cd KerasObjectDetector

  # If there is no 'setuptools' in docker, please download This package.
  # pip install setuptools
  # install library
  $ apt-get install libatlas-base-dev libxml2-dev libxslt-dev python-tk
  
  # build setup code
  # ./KerasObjectDetector
  $ python setup.py install

If you want to running on Docker, Download Docker Image, which we made, from Docker HUB and easily configure development environment.

  #  pull yolk docker image
  $ docker pull kerasyolk/yolk

  # yolk run
  $ docker run --name=yolk -Pit -p 8888:8888 -p 8022:22 kerasyolk/yolk:latest

  # running jupyter-notebook
  $ jupyter-notebook

Tutorial

You can test your image with YOLK API. Go to the Tutorial 😃

Quick Start

to be added later...

Dependencies

Name	Version(Min)
Tensorflow	1.14.0
Keras	2.3.0
Python	3.6
Numpy	1.14
Matplotlib	latest
SciPy	0.14
h5py	latest
Pillow	latest
progressbar2	latest
opencv-python	3.3.0
six	1.9.0
PyYAML	latest
Cython	latest

Release information

ver 1.0.0 (November 20, 2019)

Finally, API that can detect multiple objects in keras has been completed!! There are still many things to supply, but we plan to continue to update. This release includes:

A three of object detetion model and a data generator that changes in a suitable data format for selected model.
- Object Detection Models : SSD, YOLOv3, RetinaNet
- Dataset and data generator : PASCAL VOC2012, COCO, Custom dataset
  ㄴ Yolk's dataset downloader is 3X faster than existing downloader.
Docker files that help to set up easliy development environment.
Easy & Detail Obejct Detection Tutorial (SSD+VOC2012)

Contributors

Thanks goes to these beautiful peaple (github ID) : @fuzzythecat, @mijeongjeon, @tykimos, @SooDevv, @karl6885, @EthanJYK, @minus31, 김형섭, 최민영, @mike2ox, @hngskj, @hics33, @aaajeong, @parkjh688, @Uwonsang, @simba328, @visionNoob, 이혜리, @melonicedlatte, 전지영, @ahracho

jeonggunlee/KerasObjectDetector