- Andy Wang
Andy and I did research on computer vision and automatic game mapping. Andy applied for and won a UCSC Koret Scholarship to support our work.
- Nikolai Chen
Nikolai explored compositional approaches to solving Zelda dungeons in the HyPED modeling tool, applying search algorithms and abstraction to solve tricky problems without (much) knowledge of a game’s rules.
- Brian Lambrigger
Brian developed the most recent version of HyPED’s interactive player application, and also researched the use of rapidly exploring random trees for reachability checking in HyPED (this work led to a Best Student Paper win at AIIDE 2017).
- Freeman Tao
Freeman applied recurrent neural networks to the problem of generating hypothetical gameplay walkthroughs based on the GameFAQs text corpus.
- Andrew Gwinner
Andrew built many of the HyPED maps and models used in later research.
- Calvin Walantus
Calvin built a converter from the Videogame Level Corpus format to HyPED, and we explored possible extensions to the VGLC format in a workshop submission.
- Julius Mazer
Julius looked into generalizations of Mappy, focusing on aggregating multiple observations of the same room from different times or different playthroughs.
- Javier Alex Juanillo
Javier built several of the HyPED models used in later research and gave important feedback on the documentation.
- Desmond Vehar
Desmond explored general game playing research in the context of Gamelan, a precursor system to HyPED and other more recent work. Desmond integrated Gamelan with CADIAplayer, a game-playing program which had been successful in competition settings.