TuragaLab/flybody: MuJoCo fruit fly body model and locomotion RL tasks (github.com)

Google DeepMind and HHMI Janelia Research Campus have unveiled the "flybody," a detailed MuJoCo model simulating the locomotion of fruit flies. This innovative collaboration aims to serve as a robust platform for studying fruit fly biophysics and exploring neural control in sensorimotor behavior. The model's accessibility, demonstrated with straightforward Python integration examples, allows researchers to engage with it effortlessly, enabling tasks such as walking imitation and vision-guided flight. This development is significant for the AI/ML community as it opens up opportunities for advancing reinforcement learning (RL) techniques in biologically inspired environments. The flybody model supports various installation modes, including core functionalities for general experimentation, as well as optional ML and distributed training extensions via Ray, which cater to diverse research needs. With the integrated capabilities to simulate complex locomotion and potentially train policy networks, the flybody model is poised to enhance our understanding of both biological systems and their applications in machine learning.