Skip to content

ameasere/Railguard

Repository files navigation


Railguard
Railguard

SCR AI trained to recognise signals in realtime.

Key FeaturesHow To UseCreditsRelatedLicense

NOTICE: This project is no longer being maintained due to an internal discussion with SCR's Board of Directors. All current source code is pushed into the dev branch, which is unfinished and is simply the state the project was in when the discussion concluded. The master branch has the working version of the AI, and dev is what the project was going to be revamped to look like.

Key Features

  • Fully trained AI with templates and assets for your own use.
  • Templates for extracting training data from.
  • Administrative scripts for training the AI, testing the model accuracy etc.

How To Use

To use Railguard:

  • Source Code
    • Clone this repository via the Command Line with git clone https://github.com/ameasere/Railguard.
    • Install Python <3.10 for your OS/Architecture, and add to your PATH.
    • Install Anaconda and set up a development environment.
    • Install Python in Anaconda, preferably 3.9.
    • Install the project requirements via the requirements.txt file, make sure you have the GPU version of Tensorflow in Anaconda.

    Any issues with installation, consult this video.

    • Execute the main.py file via the Command Line with python main.py, or using an IDE.
      • We highly recommend PyCharm, that is how we developed Railguard!

Note As of the last README commit/update, Python 3.10+ is not supported. This is because Tensorflow GPU support was dropped after this version. Versions below 3.10 are tested and verified to work; the older you pick, the less likely however.

Open-Source

Please keep in mind that this software is entirely open-source, meaning everything you see was developed for free with no financial incentive, investment or gain and was entirely done during personal time.

Credits

This software uses the following open-source packages:

With a huge thanks to:

Support

Buy Me A Coffee

License

GPL-3


Developer @ameasere  ·  Operations Director @ SCR @GameZotto

About

SCR AI trained to recognise signals.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages