ID 原文 译文
24515 缓存管理已经成为机会网络的研究热点，目前大多数缓存管理方法都集中在单节点，本文在校园协作学习背景下提出了缓存共享的协作小组资源调度方法。 Cache management has become a research hot spot in opportunistic networks。 At present, most cache management methods are concentrated on single node. In this paper, we propose a cache sharing resource scheduling method for collaborative groups under the background of campus collaborative learning.
24516 节点活跃度描述了移动节点的缓存空间利用率和节点参与消息转发积极性，消息树描述了消息在网络中的扩散程度。 Node activity describes the cache spaceutilization of mobile nodes and the enthusiasm of nodes to participate in message forwarding. Message tree describes the degree of message diffusion in the network.
24517 以协作小组为单位建立已送达消息链表，对已成功投递消息所占缓存空间及时进行释放。 A linked list of delivered messages is established based on the collaboration group, and the cache space occupied by the successfully delivered messages is released in time.
24518 该方法能有效提升节点空间利用率和消息投递率并且降低了消息的投递延迟。 This method can effectively improve the node space utilization and message delivery rate, and reduce the message delivery delay.
24519 实验结果表明，与已有算法相比，消息投递成功率平均提高了10.53%，为小组缓存的共享利用与资源的高效调度提供方法。 The experimental results show that, compared with the existing algorithms, the success rate of message delivery is increased by 10.53% on average, which provides a method for the sharing and utilization of group cache and the efficient scheduling of resources.
24520 人体姿态估计是计算机视觉研究领域的热点研究问题之一，但其在传统民间舞蹈动作姿态估计方面的应用研究尚处于起步阶段。 Human pose estimation is one of the hot research topics in the field of computer vision, but its application in traditional dance pose estimation is still in its infancy.
24521 由于舞蹈图像中人体动作复杂多变、舞蹈动作连贯性强、舞蹈者存在严重遮挡不易检测等特点，传统人体姿态估计方法难以准确估计舞蹈者的动作变化，导致舞蹈动作姿态估计准确率较低。 Due to the complexity of dance pose, the strong coherence of dance movements, and difficulty in detecting of dancers' poses caused by serious occlusion in dance images, the traditional human pose estimation methods are difficult to accurately estimate the pose changes of dancers, thus resulting in low accuracy in estimating dance pose.
24522 针对此问题，本文提出一种基于序列多尺度特征融合表示的层级舞蹈动作姿态估计方法，该方法针对舞蹈动作骨骼关节点尺度变化剧烈的问题，构建基于序列多尺度特征融合表示的关节点估计模型。 We propose a hierarchical dance pose estimation method based on sequential multi-scale feature fusion. To address the problems of the drastic scale changes of the dancer pose, a keypoint estimation model based on sequential multi-scale feature fusion is constructed.
24523 并且，针对舞蹈姿态形变较大，遮挡严重的问题，设计基于关节点几何关系的层级姿态估计模型，提高舞蹈动作姿态估计的效果。 Furthermore, aiming to solve the issues that the large deformation and serious occlusion of dance pose, a hierarchical pose estimation model based on the geometric relationship between human key points is designed to improve the accuracy of dance pose estimation.
24524 实验结果表明，本文方法在标准人体姿态估计数据集及自建舞蹈数据集上取得较好的姿态估计结果。 The experimental results show that the proposed method can achieve good pose estimation results on the standard human pose estimation dataset and the self-collected dance dataset.