Boyu Chen

We present SiamSampler, the first to our knowledge investigating video sampling in visual object tracking. We observe that the random sampling applied in Siamese-based trackers cannot focus on important data or ensure data diversity, hindering the effective training of networks. This paper proposes the Video-Guided Sampling Strategy to solve the problems in random sampling from both inter and intra-video levels. At the inter-video level, we propose Modified Gaussian Sampling Strategy (MGSS) to automatically assign higher sampling probabilities to longer and more difficult videos and reduce the sampling probabilities of shorter and easier videos. At the intra-video level, the Farthest Image Pair Sampling Strategy (FPSS) is proposed to increase the diversity of training data. Extensive experiments on general benchmarks demonstrate the effectiveness of our method. Compared with the baseline model, our method improves tracking performance on five datasets, without affecting the testing speed.

Unmanned Aerial Vehicles (UAV) are widely used in different fields of aviation today. The efficient delivery of packages by drone may be one of the most promising applications of this technology. In logistic UAV missions, due to the limited capacities of power supplies, such as fuel or batteries, it is almost impossible for one unmanned vehicle to visit multiple wide areas. Thus, multiple unmanned vehicles with well-planned routes become necessary to minimize the unnecessary consumption of time, distance, and energy while carrying out the delivery missions. The aim of the present study was to develop a multiple-vehicle mission dispatch system that can automatically compile a set of optimal paths and avoid passing through no-travel zones. For this function, the A* search algorithm was adopted to determine an alternative path that does not cross the no-travel zone when the distance array is set, and an improved two-phased Tabu search was applied to converge any initial solutions into a feasible solution. In this study, a group of five multicopters was set up to validate the swarm system, and the result shows that our improved 2TS+2OPT is able to converge to a better solution that allows logistic UAV swarms to operate in a more efficient way.

The fully-convolutional siamese network based on template matching has shown great potentials in visual tracking. During testing, the template is fixed with the initial target feature and the performance totally relies on the general matching ability of the siamese network. However, this manner cannot capture the temporal variations of targets or background clutter. In this work, we propose a novel gradient-guided network to exploit the discriminative information in gradients and update the template in the siamese network through feed-forward and backward operations. Our algorithm performs feed-forward and backward operations to exploit the discriminative informaiton in gradients and capture the core attention of the target. To be specific, the algorithm can utilize the information from the gradient to update the template in the current frame. In addition, a template generalization training method is proposed to better use gradient information and avoid overfitting. To our knowledge, this work is the first attempt to exploit the information in the gradient for template update in siamese-based trackers. Extensive experiments on recent benchmarks demonstrate that our method achieves better performance than other state-of-the-art trackers.

The study aims to determine the changes in the biomechanical properties of the anterior and extreme posterior portions of experimental near-sighted eyes by examining the mechanical behavior of guinea pig scleral desmocytes, thus finding a new approach to the pathogenesis of myopia and their corresponding therapies.