Skip to content

Application of a neuroevolution algorithm to beat Touhou11. Uses automated testing tools and the NEAT genetic encoding.

Notifications You must be signed in to change notification settings

jiama843/Touhou11AIProject

Repository files navigation

Touhou11AIProject

Work in Progress using Atlassian SourceTree, will commit to Github occasionally.

Using NE to beat Touhou11. First step is to beat the first stage on hard.

Tools

  • Python
  • TouhouToolKit (commandline toolkit to extract and modify .DAT files)
  • Everything is constructed using traditional Python data structures, the Tensorflow library is not used.

May 6th 2018 update:

First commit from Atlassian includes:

  1. Population Generation
  2. Network Evaluation
  3. NodeGenes, ConnGenes and Network templates
  4. Mutation functions
  5. ScreenReader
  6. Controller Script
  7. Activation function templates

To Do: (6 items)

To Use:

  • Install Touhou 11
  • Input the right screen area
  • Run Controller.py
  • Watch it go

About

Application of a neuroevolution algorithm to beat Touhou11. Uses automated testing tools and the NEAT genetic encoding.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published