Compared with wheeled mobile robots, legged robots can easily step over obstacles and walk through rugged ground. They have more flexible bodies and therefore, can deal with complex environment. Nevertheless, some other issues make the locomotion control of legged robots a much complicated task, such as the redundant degree of freedoms and balance keeping. From literatures, locomotion control has been solved mainly based on programming mechanism. To use this method, walking trajectories for each leg and the gaits have to be designed, and the adaptability to an unknown environment cannot be guaranteed. From another aspect, studying and simulating animals' walking mechanism for engineering application is an efficient way to break the bottleneck of locomotion control for legged robots. This has attracted more and more attentions. Inspired by central pattern generator (CPG), a control method has been proved to be a successful attempt within this scope. In this paper, we will review the biological mechanism, the existence evidences, and the network properties of CPG. From the engineering perspective, we will introduce the engineering simulation of CPG, the property analysis, and the research progress of CPG inspired control method in locomotion control of legged robots. Then, in our research, we will further discuss on existing problems, hot issues, and future research directions in this field. biological inspired control, central pattern generator (CPG), locomotion control
Fig. 1. Upper two rows show projected 3D object detection results from the detector trained with binary cross entropy. Lower two rows present related results from the detector trained with the focal loss. Purple and blue bounding boxes are the ground-truth and the estimated results respectively.Abstract-3D object detection is still an open problem in autonomous driving scenes. When recognizing and localizing key objects from sparse 3D inputs, autonomous vehicles suffer from a larger continuous searching space and higher fore-background imbalance compared to image-based object detection. In this paper, we aim to solve this fore-background imbalance in 3D object detection. Inspired by the recent use of focal loss in imagebased object detection, we extend this hard-mining improvement of binary cross entropy to point-cloud-based object detection and conduct experiments to show its performance based on two different 3D detectors: 3D-FCN and VoxelNet. The evaluation results show up to 11.2AP gains through the focal loss in a wide range of hyperparameters for 3D object detection.
If citing, it is advised that you check and use the publisher's definitive version for pagination, volume/issue, and date of publication details. And where the final published version is provided on the Research Portal, if citing you are again advised to check the publisher's website for any subsequent corrections.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.