This article discusses the development of an anatomically detailed whole-body model of Drosophila melanogaster within a physics simulator. This model is designed to provide an accurate representation of the fruit fly's biomechanics to study its sensorimotor behaviors. The researchers validated the model by simulating realistic terrestrial and aerial locomotion and created models for fluid and adhesion forces. They utilized data-driven reinforcement learning to train neural network controllers for naturalistic movements. This open-source framework aids in further understanding the neural control mechanisms underlying complex behavior in the fruit fly.
We introduce a whole-body model of Drosophila melanogaster in a physics simulator that accurately represents the biomechanics underlying sensorimotor behaviors, enabling diverse locomotion simulations.
The model supports complex fly behaviors through newly developed models for fluid and adhesion forces, enhancing the realism of both terrestrial and aerial movement simulations.
Collection
[
|
...
]