myModifications.md

Project Title

This project utilizes MediaPipe's Face Landmarker to process video files and extract facial blendshape data, saving the results to a CSV file.

Modifications

1. GitHub Workflow Update

The .yaml configuration file has been updated to ensure successful application builds on GitHub.

2. Windows User Enhancements

Several issues specific to Windows users have been addressed:

• Model Path Resolution: The application now dynamically constructs the path to the face_landmarker.task model file, ensuring it's correctly located regardless of the user's environment.

• File Dialog Behavior: The file explorer dialog is now configured to appear in the foreground, preventing it from opening behind other applications.

3. Output CSV Directory Control

Users can now specify the directory for the generated CSV file, providing greater flexibility in managing output data.

Building the Executable Locally

To build the executable file locally, follow these steps:

1. Upgrade pip:

   python -m pip install --upgrade pip

2. Install flake8:

   pip install flake8

3. Install Required Packages:

   pip install -r Face_Landmarker/requirements.txt

4. Build the Application:

   • Option 1: Run the script directly:

     python Face_Landmarker/Face_Landmarker_Link.py

   • Option 2: Use PyInstaller to create an executable:

     pyinstaller Face_Landmarker/Face_Landmarker_Link.spec

     This command will generate the executable in the dist directory.

Note: Ensure that all dependencies are installed and properly configured before building the application, also you can builded it on github

For more information on creating executables with PyInstaller, refer to the PyInstaller Usage Guide.

By following these instructions, you can successfully build and run the application on your local machine.