Saber Fencing Hit Allocation with YOLO Pose

Getting Started

To run this pipeline, follow these steps:

pip install -r requirements.txt
python main.py # run the pipeline

Repo Overview

Here is an overview of the important parts of the project:

• main.py: Runs the mainloop of the fencing scoring pipeline
  • -m, --mode: Inference mode [optional]; either "webcam" or "file"
  • -f, --file: Path to the file you want to run inference on [optional]
  • --headless: Use if you are running in a container or with opencv-python-headless
• fencer_pose.py: Holds the Class for our Fencing Pose Estimator
• scorebox_classifier.py: Holds the Class for detecting points based on the fencing scorebox
• yolo_scorebox_classifier.py: Holds the Class for both detecting scoreboxes as well as detecting points based on the detected scorebox
• nn_pose_classifier.py: Holds the Class for reading pose estimations and classifying which action they correspond to
• classifier_data.py: Holds functions to take pose classification labels and convert them to a readable format