Learning Transferable UAV for Forest Visual Perception
This work addresses the challenge of UAV navigation in complex forest environments, but it is incremental as it builds on existing methods like ResNet-18 and domain adaptation.
The paper tackles the problem of training a UAV to fly collision-free along dense forest trails by proposing a pipeline that uses a simulated dataset and a ResNet-18 model for direction classification, achieving 84.08% accuracy in real-world tests.
In this paper, we propose a new pipeline of training a monocular UAV to fly a collision-free trajectory along the dense forest trail. As gathering high-precision images in the real world is expensive and the off-the-shelf dataset has some deficiencies, we collect a new dense forest trail dataset in a variety of simulated environment in Unreal Engine. Then we formulate visual perception of forests as a classification problem. A ResNet-18 model is trained to decide the moving direction frame by frame. To transfer the learned strategy to the real world, we construct a ResNet-18 adaptation model via multi-kernel maximum mean discrepancies to leverage the relevant labelled data and alleviate the discrepancy between simulated and real environment. Simulation and real-world flight with a variety of appearance and environment changes are both tested. The ResNet-18 adaptation and its variant model achieve the best result of 84.08% accuracy in reality.