This repo comprises of the code to run a detection algoithm on a server as a service and it has live implementation of my nosave-server approach. Use it freely :)
Highly recommended to have Tensorflow-gpu up and running.
After moving inside the directory. To run the project:
- Install virtual environment
pip install virtualenv- Initiate virtual environment
virtualenv venv- Activate virtualenv
. venv/bin/activate- Install the requirements
pip install -r requirements.txt- Download and convert the weights for yolov3 and yolov3-tiny
wget https://pjreddie.com/media/files/yolov3.weights
./yad2k.py yolov3.cfg yolov3.weights yolo.h5
wget https://pjreddie.com/media/files/yolov3-tiny.weights
./yad2k.py yolov3-tiny.cfg yolov3-tiny.weights tint_yolo.h5- Run the server
python server.py