Skip to content

README.md

Latest commit

 

History

History
19 lines (15 loc) · 902 Bytes

README.md

File metadata and controls

19 lines (15 loc) · 902 Bytes
Raw

Action Recognition

This repository contains the code and models for an action recognition system that classifies actions into three categories: Boxing, Kick, and Nonfighting. The system is part of a pipeline to integrate action recognition into Unity using Unity Sentis.

Git for Unity project related to the pipeline: https://github.com/ArianMoeini/UnityDone

Repository Structure

  • DifferentSavedData/: Directory for various saved datasets.
  • Logs/: Directory for storing log files.
  • MP_Data/: Directory for storing MediaPipe data.
  • ActionDetection.ipynb: Jupyter notebook containing the main code for action detection.
  • README.md: This file.
  • extraData.zip: Additional data for the project.
  • mainModel.h5: The main model file in H5 format.
  • model.onnx: The model file in ONNX format.
  • model2.h5: Another model file in H5 format.