Uma biblioteca para a criação de jogadores virtuais em emuladores

Abstract

Among the areas studied in Computer Science, Artificial Intelligence has recently stood out not only in the academic community but also in industry, especially in the area of games. Considering that there is demand for artificial intelligence environments to create and test new technologies, the present bachelor thesis has the purpose of defining and building a software library that will serve as an interface for the creation and testing of algorithms of artificial intelligence with the Multiple Machine Arcade Emulator (MAME), which is the emulator with more games available and greater variety of hardware architectures, since it emulates several different platforms, in which each game can serve as a new environment, with its own rules. In order to achieve this goal, a Python library has been developed that can interact with MAME games, offering the possibility to use the game frames and send keyboard events to the emulator to interact with the game being played. This work contributes with the definition and implementation of a Python library that can be used by anyone who wants to built artificial intelligence algorithms for MAME emulator games.