High-level deep learning package for Object Detection API


Keywords
augmentation, image, deep, learning, neural, network, CNN, machine, computer, vision, object, detection, api, tensorflow, data, preprocessing, manipulation, visualization
License
MIT
Install
pip install vebits-api==1.1.2

Documentation

vebits_api

A high-level, comprehensive package that leverages user's experience when working with Tensorflow's Object Detection API.

Overview

This package has been developed to turn my works at Vebits into a friendly, easy-to-use API that facilitate user's experience when working with Tensorflow Object Detection API. New features are being developed and tested to working with DarkNet/Darkflow for training YOLO models running real-time on mobile devices; as well as with MMdetection for high-performance, highly scalable object detection toolbox.

Dependencies

All dependencies are listed under requirement.txt

certifi==2019.6.16
cycler==0.10.0
decorator==4.4.0
imageio==2.5.0
imgaug==0.2.9
imutils==0.5.2
kiwisolver==1.1.0
matplotlib==3.1.1
networkx==2.3
numpy==1.16.4
opencv-python==4.1.0.25
pandas==0.25.0
Pillow==6.1.0
protobuf==3.9.0
pyparsing==2.4.1
python-dateutil==2.8.0
pytz==2019.1
PyWavelets==1.0.3
scikit-image==0.15.0
scipy==1.3.0
Shapely==1.6.4.post2
six==1.12.0
tqdm==4.32.2

Optionally, the following packages are required for the API to work seamlessly with

  • Tensorflow's Object Detection API: tensorflow
pip install tensorflow-gpu
  • Darknet/Darkflow for YOLO models: darkflow
git clone https://github.com/thtrieu/darkflow.git
cd darkflow
pip install -e .
  • MMdetection toolbox: details on installation can be found here.

Installation

To install the latest stable release of this package, simply run:

pip install vebits_api

Alternatively, to build the project from source in development mode and allow the changes to take effect immediately:

git clone https://github.com/hnt4499/vebits_api/
cd vebits_api/
pip install -e .

or

pip install git+https://github.com/hnt4499/vebits_api.git

That's it! To make use of available scripts for data manipulating/processing/visualization, simply copy all scripts under scripts folder to your working directory.

TODO:

  • Complete README.md: Requirements, Build from source, Usage, Reference, Examples.
  • Incorporate DarkNet into this package.
  • [ ]