Person and head extraction with green screen removal

Installation

1. Clone this repository

   git clone https://github.com/adamp87/green-screen-extractor.git;

2. Go into the repository

   cd green-screen-extractor;

3. Create conda environment and activate

   conda create -n gse python=3.8,

   conda activate gse;

4. Install dependencies

   pip install -r requirements.txt,

5. Download pretrained model (non-commercial)

   Link from repository by deepakcrk: yolov5-crowdhuman

   Download Link: YOLOv5m-crowd-human

Licensing

Source code uses YOLO5, which is licensed under GPLv3.

Pretrained model is trained on the crowdhuman dataset, which is non-commercial. CrowdHuman