Fly360: Omnidirectional Obstacle Avoidance within Drone View
This work tackles the problem of omnidirectional obstacle avoidance for UAVs, which is crucial for safe and robust drone operation in complex environments, offering an incremental improvement over existing limited field-of-view methods.
This paper addresses the problem of omnidirectional obstacle avoidance for drones, where movement direction can differ from the drone's heading. The authors propose Fly360, a two-stage perception-decision pipeline that uses panoramic RGB observations converted to depth maps, enabling stable omnidirectional obstacle avoidance and outperforming forward-view baselines across all tasks in simulations and real-world experiments.
Obstacle avoidance in unmanned aerial vehicles (UAVs), as a fundamental capability, has gained increasing attention with the growing focus on spatial intelligence. However, current obstacle-avoidance methods mainly depend on limited field-of-view sensors and are ill-suited for UAV scenarios which require full-spatial awareness when the movement direction differs from the UAV's heading. This limitation motivates us to explore omnidirectional obstacle avoidance for panoramic drones with full-view perception. We first study an under explored problem setting in which a UAV must generate collision-free motion in environments with obstacles from arbitrary directions, and then construct a benchmark that consists of three representative flight tasks. Based on such settings, we propose Fly360, a two-stage perception-decision pipeline with a fixed random-yaw training strategy. At the perception stage, panoramic RGB observations are input and converted into depth maps as a robust intermediate representation. For the policy network, it is lightweight and used to output body-frame velocity commands from depth inputs. Extensive simulation and real-world experiments demonstrate that Fly360 achieves stable omnidirectional obstacle avoidance and outperforms forward-view baselines across all tasks. Our model is available at https://zxkai.github.io/fly360/