MimiMori is a game made as a team during Scientific Game Jam Online 2026. The game is to familiarize the public with the way deeplearning models are used for bioacoustics researches. These algorithm has to be lightweight in order to fit in technically limited devices, while being efficient enough to record and identify only the sounds that are useful for researchers. This theme was the work subject of our team scientist member, Alexandre Bento, that was supervising the way we handled the scientific aspects in game.
The team : Alexandre Bento (Scientist), Andry Rajanamavo (UI Designer), Clara Laborderie (Game Designer), Edwin Miko (Sound Designer), Johanna Birraldacci (Artist), Julien Lemaire / Sticmac (Developer), Quentin Chauvineau (Sound Designer), XLA (Developer)
We choose to give players the role of the algorithm that has to take choices in data extraction to help them understood the purpose of it whereas of treating it only like a mechanic.
In game this result in having to select audio extracts from the spectrogram before sending them to the research team. However, each thing that get sent cost some battery to the device running the algorithm. So to survive the mission, players have to pick only what is relevant, else they might end up out of battery or worse, get reprogrammed if what they've sent was not up to researchers standards.
Itch page (only available in french) : https://claratosaure.itch.io/sgj-2026-mimi-mori
Tools and software's
Project gallery
