Search

It is necessary for the development of self-driving to fulfill the requirements for safety, stability, and intelligence, especially in high-speed conditions. Therefore, the detection of pedestrians that may occur in highway scenarios during driving and understanding the meaning of their behaviors in advance are significantly important for the self-driving vehicle to make correct decisions. However, no existing datasets are available for behavior recognition in self-driving scenarios. In order to advance the task of interactive cognition between the vehicle and pedestrians, in this paper, we present a new dataset, called THU-IntrudBehavior, that collects lane intrusion behaviors of pedestrians that can be applied in real world highway scenarios. The dataset contains diverse behaviors of single or multiple pedestrians/cyclists that are simulated in different urban roads under various weather conditions. We describe annotations of each video and report several experimental results of baseline methods on our self-collected dataset. Our THU-IntrudBehavior dataset provides new support for behavior recognition in high-speed conditions for self-driving.

In a complex road traffic scene, illegal lane intrusion of pedestrians or cyclists constitutes one of the main safety challenges in autonomous driving application. In this paper, we propose a novel object-level phase space reconstruction network (PSRNet) for motion time series classification, aiming to recognize lane intrusion actions that occur 150m ahead through a monocular camera fixed on moving vehicle. In the PSRNet, the movement of pedestrians and cyclists, specifically viewed as an observable object-level dynamic process, can be reconstructed as trajectories of state vectors in a latent phase space and further characterized by a learnable Lyapunov exponent-like classifier that indicates discrimination in terms of average exponential divergence of state trajectories. Additionally, in order to first transform video inputs into one-dimensional motion time series of each object, a lane width normalization based on visual object tracking-by-detection is presented. Extensive experiments are conducted on the THU-IntrudBehavior dataset collected from real urban roads. The results show that our PSRNet could reach the best accuracy of 98.0%, which remarkably exceeds existing action recognition approaches by more than 30%.

Quantization-based methods are widely used in LiDAR points 3D object detection for its efficiency in extracting context information. Unlike image where the context information is distributed evenly over the object, most LiDAR points are distributed along the object boundary, which means the boundary features are more critical in LiDAR points 3D detection. However, quantization inevitably introduces ambiguity during both the training and inference stages. To alleviate this problem, we propose a one-stage and voting-based 3D detector, named Point-Voxel-Grid Network (PVGNet). In particular, PVGNet extracts point, voxel and grid-level features in a unified backbone architecture and produces point-wise fusion features. It segments Li-DAR points into foreground and background, predicts a 3D bounding box for each foreground point, and performs group voting to get the final detection results. Moreover, we observe that instance-level point imbalance due to occlusion and observation distance also degrades the detection performance. A novel instance-aware focal loss is proposed to alleviate this problem and further improve the detection ability. We conduct experiments on the KITTI and Waymo datasets. Our proposed PVGNet outperforms previous state-of-the-art methods and ranks at the top of KITTI 3D/BEV detection leaderboards.

Planar magnetic devices have played an important role in many applications to increase performance because of the advantages it bears such as low profile and ease of manufacture. Meanwhile litz wire is a solution to improve it further by reducing ac resistance. Studies of printed circuit board (PCB) litz wire structure on planar air-core inductor and planar transformer are presented in this paper along with several theoretical analyses of resistance and inductance of PCB litz wire. Also, a novel litz wire planar transformer and 2 designs of litz wire air-core inductors are presented, simulated and tested, which shows promising results compared to conventional solid or parallel wire designs for high operating frequency operations.

A medium-frequency transformer isolated DC/DC converter with series L-C and resonant L-C shunt filter is designed and simulated in this paper to eliminate high order harmonics of square wave and therefore reduce core loss of medium-frequency planar transformer. By adding an air-core series inductor with a series resonant capacitor and a resonant shunt branch consisting of another air-core inductor in series with the capacitor, specific order of harmonics can be entirely filtered without increasing extra copper loss from the inductors. At the same time the wide voltage ratio between output and input can be controlled at the first stage due to the passiveness of the inverter and rectifier. This proposed method can decrease around 5% of core loss of transformer compared with square wave and the circuit works with only 1.90% total harmonic distortion. In the meantime, it simplified the control and has the potential to operate at several hundred kilohertz to achieve compactness and miniaturization.

© 2018 IEEE. As piracy presents great threats to maritime security, unmanned aerial vehicle (UAV) swarm monitoring is becoming an attractive countermeasure. Conventional sweep monitoring method provides balanced coverage but cannot adapt to variable situations, such as the return of UAVs for refueling. This paper proposes a pheromone-based method for UAV swarm anti-piracy monitoring. In this method, the environment is modeled by a pheromone map, and the motion of UAVs is guided by the strength of pheromones. A ship-centered convening mechanism is proposed to increase the chance of finding pirates by taking the interactions between pirates and merchant ships into account. A prediction-reservation mechanism is proposed to improve the deficiencies of the implicit intention propagation mechanism when the ship-centered convening mechanism is used. Simulation experiments are conducted to study the effectiveness of the proposed method. Results show that the proposed method reduces the success rate of pirate attack by 8% at most compared to the sweep method, and that the prediction-reservation mechanism performs better than the implicit intention propagation mechanism, especially with large number of UAVs. ispartof: pages:2253-2258 ispartof: 2018 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND BIOMIMETICS (ROBIO) pages:2253-2258 ispartof: IEEE International Conference on Robotics and Biomimetics (ROBIO) location:MALAYSIA, Kuala Lumpur date:12 Dec - 15 Dec 2018 status: published

In dynamic and competitive times, retaining strategic flexibility through strategic changing is an effective method for enterprise to cope with changes. Strategy change, like ldquooptionrdquo in the financial economics, is to gain the highest possible risk investment return. So the thought of option is introduced to give a hand to managers to gain strategic flexibility. As a quantitative analysis tool, real option is of tremendous potential value in the field of strategic management. The analysis of the relationship between option thought and strategic management is made, and the function of strategic option is discussed in this paper. Furthermore, this paper tries to apply the thought of option to strategic change and structure a strategic change model combined with option thought. The model displays how strategy change is dependent on options to spread in different ways, and displays how to build a bridge between the present and the future.