Dynamic Level Generation for Games
Contains code for Dynamic Level Generation for Games, USC CSCI 527 Fall 2020.
Github link: https://github.com/CS527Applied-Machine-Learning-for-Games/Dynamic-level-Generation-for-Games
Objective
When playing games, players often begin with different skill levels and gradually develop their skills. Game designers often face the challenge of creating a perfectly designed game which is compatible to everyone’s skill level. In tackling this, we present a proposed solution. Dynamic Level Generation is a general category of approaches focusing on basic parameter tweaking, which, in response to player performance, alters games during play. Challenging aspects influence the difficulty of many games such as level design which are difficult to adjust dynamically. Also, designer intuition is a driving force behind Dynamic Level Generation techniques, and this may not mirror actual play patterns. Since the changes are both structural and personalized, we believe Dynamic Level Generation will create a unique play experience. Along with that, this will also pose as a fitting example of a favorable new R&D approach.
The following video shows the demo of our project:
Here is the link to the presentation:
Contributors
Ayush Singh
Himanshu Singh
Karthik Sagar
Shabbir Habib
Shagun Bhatia
Shubhankar Singh
Vikas Nyamati