The Video Game Level Corpus (VGLC)
收藏资源简介:
The Video Game Level Corpus (VGLC) 是一个包含428个来自12款游戏关卡的数据集,由加州大学圣克鲁兹分校和德雷塞尔大学创建。数据集中的关卡以易于解析的文本文件形式存在,并附有相应的图像表示,适合机器学习和游戏AI研究。VGLC包含多种游戏类型,如2D平台游戏、射击游戏等,支持Tile、Graph和Vector三种注释格式。创建过程中,研究者们对游戏关卡进行了详细的解析和标注。该数据集主要用于自动生成游戏关卡的研究,旨在通过机器学习技术解决关卡设计的自动化问题。
The Video Game Level Corpus (VGLC) is a dataset consisting of 428 game levels from 12 different games, developed by researchers from the University of California, Santa Cruz and Drexel University. The levels in the dataset are stored in easily parsable text files, paired with corresponding visual representations, making it well-suited for machine learning and game AI research. VGLC covers a wide range of game genres, including 2D platformers, shooters, and more, and supports three annotation formats: Tile, Graph, and Vector. During its construction, the researchers conducted detailed parsing and annotation work on all included game levels. This dataset is primarily used for research on automatic game level generation, with the objective of addressing the automation of level design through machine learning techniques.

- 1The VGLC: The Video Game Level Corpus加州大学圣克鲁兹分校, 德雷塞尔大学 · 2016年



