Learning Games from Videos Guided by Descriptive Complexity
By Lukasz Kaiser. 2012
31 citations recorded by the source corpus.
In recent years, several systems have been proposed that learn the rules of a simple card or board game solely from visual demonstration. These systems were constructed for specific games and rely on substantial background knowledge. We introduce a general system for learning board game rules from videos and demonstrate it on several well-known games. The presented algorithm requires only a few demonstrations and minimal background knowledge, and, having learned the rules, automatically derives position evaluation functions and can play the learned games competitively. Our main technique is based on descriptive complexity, i.e. the logical means necessary to define a set of interest. We compute formulas defining allowed moves and final positions in a game in different logics and select the most adequate ones. We show that this method is well-suited for board games and there is strong theoretical evidence that it will generalize to other problems.
Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Matthew J. Hausknecht
- Joel Lehman
- Peter Stone
- J. Urgen Schmidhuber
- Kenneth O. Stanley
- Shimon Whiteson
- Jeff Clune
- Sergey Levine
In Computer Vision and Pattern Recognition
- Joel Lehman
- Matthew J. Hausknecht
- Peter Stone
- Masataro Asai
- Risto Miikkulainen
- Alex S. Fukunaga
- J. Urgen Schmidhuber
- Luc Van Gool
Citations in the corpus
- A Neuroevolution Approach to General Atari Game Playing (2014, background): “…exclusively by playing the game; 3) while the GGP competition is based on declarative representations , this paper focuses on visual representations (as suggested by other recent work in GGP [1], [22]); and 4) while GGP features board-like games, this paper takes a step toward playing video games.”
- Classical Planning in Deep Latent Space: Bridging the Subsymbolic-Symbolic Boundary (2017, background): “A closely related line of work in LfO is learning of board game play from observation of video/images [Barbu et al. , 2010; Kaiser, 2012; Kirk and Laird, 2016].”