README.md

FootAndBall: Integrated player and ball detector

Input

This image is a part of input.mp4 and input.mp4 was created from the video in ISSIA-CNR Soccer dataset

Output

Usage

Automatically downloads the onnx and prototxt files on the first run. It is necessary to be connected to the Internet while downloading.

For the sample image,

$ python3 footandball.py

You can choose predicting mode in ailia and onnxruntime. The default setting is ailia and you can use onnxruntime by adding --onnx option as below.

$ python3 footandball.py --onnx

If you want to specify the input image, put the image path after the --input option.
You can use --savepath option to change the name of the output file to save.

$ python3 footandball.py --input IMAGE_PATH --savepath SAVE_IMAGE_PATH

By adding the --video option, you can input the video.
If you pass 0 as an argument to VIDEO_PATH, you can use the webcam input instead of the video file.

$ python3 footandball.py --video VIDEO_PATH

You can customize detection threshold settings by adding --player_threshold and --ball_threshold options. The default threshold is 0.7 in both.

$ python3 footandball.py --player_threshold PLAYER_THRESHOLD

$ python3 footandball.py --ball_threshold BALL_THRESHOLD

Reference

FootAndBall: Integrated player and ball detector

Framework

Pytorch

Model Format

ONNX opset = 11

Netron

footandball.onnx.prototxt