INTRODUCTION

General python yolo framework, warp of pyyolo, including training and inference processes.

CREATE VIRTUAL ENV

create python3 virtual env:

virtualenv -p python3 --no-site-packages venv

activate virtual env:

source venv/bin/activate

INSTALL requirements

pip install -r requirements.txt

INSTALL pyyolo

clone project:

git clone --recursive https://github.com/thomaspark-pkj/pyyolo.git

Delete line 170 in module.c, then follow pyyolo README.md

PREPARATION

cd gpyyolo
cp conf.json.tmpl conf.json

make sure that model weight file is in dir: gpyyolo/

RUN

INFERENCE

cd gpyyolo
python gpyyolo.py -task inference

TRAIN

label images: tool link

generate labels:

cd gpyyolo
python data_label.py  # update its class name, data path etc. before running

compile darknet framework (todo: auto compile)

cd pyyolo/darknet
make

before running, pls check cfg params. run:

cd gpyyolo
python gpyyolo.py -task train