PPP

Privacy preserving PropTech.

Poster

Test Code

There are two versions. One is with yolov5 and LaMa Inpainting, and the other is with mask_RCNN with LaMa Inpainting. You could try it on Colab.

How to use

Yolov5+LaMa Inpainting Version

1. Use code block named "settings for ppp" for setting.

   • Git clone PPP
   • Install packages for yolov5 and LaMa Inpainting
   • Install an LaMa model file

   PPP
   │ lama   
   │  └── big-lama   
   │      ├── config.yaml # model configs  
   │      └── models           
   │          └── best.ckpt # model weights     
   

   • Make an directory for LaMa Inpainting prediction

   PPP
   │ lama
   │  └── data_for_prediction
   │      ├── sample1.png # example
   │      └── sample1_mask.png # example
   

2. Use code block named "Change file path and start object detection and Inpainting" for object detection, mask generating, and mask inpainting.

   • By using maskGenerator.yolov5Mask, you could make an mask(black and white) with yolov5.
   • Make an Inpainted Image from the mask made by the maskGenerator.yolov5Mask.

MaskRCNN+LaMa Inpainting Version

1. Use code block named "Change python version 3.10 -> 3.7" for change python version.
   • Python version change in google colab
2. Use code bloack named "Settings for ppp" for setting
   • Git clone PPP
   • Install packages for mask rcnn and LaMa Inpainting
   • Install LaMa model file

   .root
   │ lama   
   │  └── big-lama   
   │      ├── config.yaml # model configs     
   │      └── models           
   │          └── best.ckpt # model weights    
   

   • Make an directory for LaMa Inpainting prediction.

   PPP
   │ lama
   │  └── data_for_prediction
   │      ├── sample8.png # example
   │      └── sample8_mask.png # example
   

3. Use code block named "Change file path and start Inpainting" for object detection, mask generating, and mask inpainting.
   • By using maskGenerator.mrcnnMask, you could make an mask(black and white) with mask RCNN.
   • After making mask, reset gpu ram for LaMa Inpainting
   • Make an Inpainted Image from the mask made by the maskGenerator.mrcnnMask

Requirements

We changed a few codes of yolov5 and Mask RCNN for Running both models(yolov5+LaMa or mask RCNN+Lama).

Yolov5+LaMa Version

python 3.10 torch 2.0.0 cu118 etc

Mask RCNN+LaMa Version

python 3.7 torch 1.8.0 cu111 etc

License

YOLOv5 is available under two different licenses:

• AGPL-3.0 License: See LICENSE file for details.
• Enterprise License: Provides greater flexibility for commercial product development without the open-source requirements of AGPL-3.0. Typical use cases are embedding Ultralytics software and AI models in commercial products and applications. Request an Enterprise License at Ultralytics Licensing.

Citation

Use this bibtex to cite this repository:

@misc{matterport_maskrcnn_2017,
  title={Mask R-CNN for object detection and instance segmentation on Keras and TensorFlow},
  author={Waleed Abdulla},
  year={2017},
  publisher={Github},
  journal={GitHub repository},
  howpublished={\url{https://github.com/matterport/Mask_RCNN}},
}