CCExtractor / Rekognition

Free and Open Source alternative to Amazon's Rekognition service. CCExtractor Development | Poor Man's Rekognition
GNU General Public License v3.0
101 stars 52 forks source link
computer-vision deep-learning django django-rest-framework docker face-detection gsoc image-processing machine-learning machinelearning opencv python rest rest-api tensorflow tensorflow-serving video-processing

Poor Man's Rekognition


Google Summer Of Code Project under CCExtractor Development

Build Status Python 3.X GPLv3 license


This project aims at providing a free alternative to Amazon Rekognition services.

Setup

For End-User

git clone https://github.com/pymit/Rekognition

docker image build ./

Note down the IMAGEID at the end and run the docker

docker run -p 8000:8000 <IMAGEID>

For Developers

To setup the project locally for development environment check this wiki link

Usage

This project currently supports Feature cURL
Face Recognition with FaceNet curl -i -X POST -H "Content-Type: multipart/form-data " -F "file=@<path to image file> " --form network=1 http://127.0.0.1:8000/api/image/
Face Recognition with RetinaNet curl -i -X POST -H "Content-Type: multipart/form-data " -F "file=@<path to image file> " --form network=2 http://127.0.0.1:8000/api/image/
Similar Face Search curl -i -X POST -H "Content-Type: multipart/form-data" -F "file=@ <path to reference image>" -F "compareImage=@ <path to compare Image>" http://127.0.0.1:8000/api/simface/
NSFW Classifier curl -i -X POST -H "Content-Type: multipart/form-data " -F "file=@<path to image file> " http://127.0.0.1:8000/api/nsfw/
Text Extraction curl -i -X POST -H "Content-Type: multipart/form-data " -F "file=@<path to image file> " http://127.0.0.1:8000/api/scenetext/
Object Detection curl -i -X POST -H "Content-Type: multipart/form-data " -F "file=@<path to image file> " http://127.0.0.1:8000/api/objects/
Scene Classification curl -i -X POST -H "Content-Type: multipart/form-data " -F "file=@<path to image file> " http://127.0.0.1:8000/api/scenedetect/

Details on documentation can be found here.

Communication

Real-time communication for this project happens on slack channel of CCExtractor Development, channel link. You may join this channel via this link

References

This project uses the following.

  1. FaceNet
  2. CRNN
  3. EAST
  4. Synth90k
  5. YOLOv3
  6. Places365
  7. RetinaFace

License

This software is licensed under GNU GPLv3. Please see the included License file.