ID 原文 译文
39046 OSNet是一种有效的轻量级行人重识别网络，因其兼具有轻量化和高性能的优异特点引起了行人重识别领域的关注。 OSNet is an effective lightweight neural network architecture, which has attracted attention in the field of person re-identification duo to its excellent performance.
39047 最近的研究表明:多分支协作OSNet网络——BC-OSNet能取得更高的识别率。 Recently, we have proposed a multi-branch cooperative OSNet network based on OSNet, termed BC-OSNet, which performs significantly better than OSNet.
39048 本文在此基础上继续研究网络微结构的调整对BC-OSNet模型性能的影响，重点通过通用池化GeM、连续高斯Dropout、注意力学习Batch DropBlock(BDB)/Relation-Aware Global Attention (RGA)等微结构的有效融入，研究微结构优化的BC-OSNet性能提升效果。 In this paper, we study the further optimization of BC-OSNet with some adjustments on various micro-structures, including generalized-mean pooling, continuous Gaussian Dropout, attention modules of Batch DropBlock(BDB)/Relation-Aware Global Attention(RGA), etc.
39049 实验结果表明:经微结构优化的BC-OSNet在四个行人重识别数据集Market1501，Duke，CUHK03Labeled和CUHK03Detected上的mAP分别达到了89.9%，82.1%，84.2%和81.5%，相比初始的BC-OSNet提高0.6%，1.4%，1.1%和1.7%。 Experimental results show that the optimized BC-OSNet achieves 89.9%, 82.1%, 84.2%, and 81.5% mAP on the four pedestrian re-identification datasets, including Market1501, Duke, CUHK03Labeled, and CUHK03Detected, respectively. This means that the optimized BC-OSNet surpasses BC-OSNet about 0.6%, 1.4%, 1.1% and 1.7% in mAP for these datasets.
39050 在传感器网络的多目标跟踪研究中，大多数现有的跟踪算法通常设定网络中所有节点具有相同的视野，即所有节点都能够得到目标的测量，但在实际中，节点的感测范围通常是有限的。 In the study of multi-target tracking in sensor networks, most of the existing tracking algorithms generally assume that all nodes in the network have the same field of view, that is, all nodes can obtain the target measurement. But in practice, the sensing range of nodes is usually limited.
39051 针对这一问题，本文提出一种能够在感测范围有限的多传感器网络中实现多目标跟踪的分布式概率假设密度滤波算法， To solve this problem, we proposes a distributed probability hypothesis density filtering algorithm that can realize multi-target tracking in sensor networks with limited sensing range.
39052 该算法通过融合传感器网络视野范围内的后验概率假设密度粒子集来克服传感器节点感测范围的局限。 This algorithm overcomes the limitation of the sensing range of sensor nodes by fusing the particle set of posterior probability hypothesis density in the field of view of the sensor network.
39053 仿真结果表明，提出的算法可以在感测范围有限的情况下实现多目标状态和数目的有效跟踪，同时可以在一定程度上抑制杂波，具有较好的跟踪稳定性。 The simulation results show that the proposed algorithm can not only achieve effective tracking of multiple target states and numbers with limited sensing range, but also have a certain effect of clutter suppression, and have good tracking stability.
39054 在未来物联城市的规划和现有的发展情况下，基于位置的服务变得越来越重要。 Location-based services can be used to a great effect in city planning and the development of the future IoT cities.
39055 室内定位在很多场景都有着明朗的应用前景，如轨迹跟踪、老人看护、手势识别等。 Device-free passive indoor localization is playing a critical role in many applications such as tracking, elderly care, gesture recognition, etc.