Path Stitching: Controllable Racing Path Generation

摘要：

In this paper we propose an alternative way of producing racing path data. Some applications require not only fast path generation time, but also restrict the finish rankings of the path. Using a data-driven method, we search the path database for dips that can be joined together in the preprocessing phase. In the application, we combine the existing paths in the database and produce a new racing path that satisfy the rankings requirement on-the-fly. Our method is fast and capable of producing good quality paths because only good quality paths are included in the path database.

关键词： AI tool Racing game Data-driven path generation Race course

作者: CHARLIE IRAWAN TAN WEN-KAI TAI

作者单位: Departement of Computer Science and Information Engineering, National Dong Hwa University 97401 Da Hsueh Rd.Hualien, Taiwan R.O.C.

会议类型: 国际会议

会议名称: 2009 International Conference on Machine Learning and Cybernetics(2009机器学习与控制论国际会议)

会议地点: 保定

会议语种:英文

页码: 2939-2944

在线出版日期: 2009-07-12（万方平台首次上网日期，不代表论文的发表时间）