Object detectors have come a long way and are used for various applications. In pictures and videos, an object detector must deal with the background. In some settings, this background is indicative of the object; in others, it’s not and can even be disruptive. For models trained on data containing correlations between objects and backgrounds (background bias), it makes sense that changing the background can disrupt learned correlations. This paper is interested in how sensitive object detectors are to background changes, specifically when the training data does not contain correlations between objects and backgrounds. Models were trained on carefully controlled synthetic data, so only the backgrounds differed and correlations could be controlled. The results show that models perform better when tested with seen backgrounds than unseen backgrounds. This performance difference diminishes when the model is trained on more unique backgrounds.

Raccoon Serious Games develops different kinds of gaming experiences, including escape rooms. In an escape room, a group of players, usually between 2 and 20 people, are locked in a room, where they have to find clues and solve puzzles to escape. When such a room is played, there is always an operator, monitoring the progress of the players and keeping an eye on their safety. Modern escape rooms are quite technologically advanced, where all components of the room interact with each other. To run these games, Raccoon Serious Games needs a system which should manage all technical aspects within the room and provide a way for the operator to keep track of the players. Furthermore, the system must be flexible enough to handle many different escape rooms, which are hardcoded in configuration files. To meet these needs of Raccoon Serious Games, a team of three developers has developed a new system, called S.C.I.L.E.R.. In ten weeks, they have created a system from scratch. First, they researched technologies and requirements that would be necessary for the system and afterwards implemented the system based on their findings. S.C.I.L.E.R. allows an operator to monitor a complete escape room. The system connects a user interface to all the devices in the room, which are controlled by the user as well as the configuration of the escape room. This configuration is generated from a JSON configuration file, containing all information for the escape room. The system allows the operator to send hints to the players, control the time and manage their progress through the status of devices in the room and the puzzles they have to solve. It also contains the feeds of cameras in the room to actually see the players. Furthermore, it provides the user with a way to check a configuration file and put it to use. The system will be used by Raccoon Serious Games in the near future for the escape rooms that they will be hosting.