ID 原文 译文
663 同时，采用基于整体奖励信号优化策略，相比于使用损失函数启发式优化特定的单个决策，该方法直接优化整体评估指标更加高效。 In this study, we use an overall reward signal optimization strategy, which is more efficient than directly using the loss function heuristicto optimize a specific single decision.
664 最后在维吾尔语数据集进行实验。 Finally, we conduct experiments in the Uyghur dataset.
665 实验结果显示，该方法在维吾尔语人称代词指代消解任务中的 F 值为 85.80% 。 The experimental results show that the F value of this method in the Uyghur personal pronouns resolution task is 85. 80% .
666 实验结果表明，深度强化学习模型能显著提升维吾尔语人称代词指代消解性能。 The experimental results show that the deep reinforcement learning model can significantly improve the performance of the Uyghur personal pro-nouns resolution.
667 目标检测是计算机视觉领域内的热点课题，在机器人导航、智能视频监控及航天航空等领域都有广泛的应用。 Object detection is a hot topic in the field of computer vision, and has been widely used in robot naviga-tion, intelligent video surveillance, aerospace, and other fields.
668 本文首先综述了目标检测的研究背景、意义及难点，接着对基于深度学习目标检测算法的两大类进行综述，即基于候选区域和基于回归算法。 The research background, significance and challenges of object detection were introduced. Then the object detection algorithms based on deep learning were reviewed according to two cate-gories:candidate region-based and regression-based.
669 对于第一类算法，先介绍了基于区域的卷积神经网络(Region with Convolutional NeuralNetwork，R-CNN)系列算法，然后从四个维度综述了研究者在 R-CNN 系列算法基础上所做的研究:对特征提取网络的改进研究、对感兴趣区域池化层的改进研究、对区域提取网络的改进研究、对非极大值抑制算法的改进研究。 For the candidate region-based algorithms, we first introduced the R-CNN (Region with Convolutional Neural Network)based series of algorithms, and then the R-CNN based methods were o-verviewed from four dimensions:the research of feature extraction networks, the region of interesting pooling researches, im-proved works based on region proposal networks, and some improved approaches of non maximum suppression algorithms.
670 对第二类算法分为 YOLO(You Only Look Once)系列、SSD(Single Shot multibox Detector)算法及其改进研究进行综述。 Next, the regression-based algorithms were surveyed in terms of YOLO(You Only Look Once)series and SSD(Single Shotmultibox Detector)series.
671 最后根据当前目标检测算法在发展更高效合理的检测框架的趋势下，展望了目标检测算法未来在无监督和未知类别物体检测方向的研究热点。 Finally, according to the current trend of object detection algorithms that are developing more effi-cient and reasonable detection frameworks, the future research focuses of unsupervised and unknown category object detec-tion directions were prospected.
672 针对多波束卫星通信系统星上资源稀缺和能量利用效率不高的问题，本文提出了分布式星群网络下行链路中兼顾系统功耗和数据速率的功率分配方法，通过合理的资源分配来优化系统的能量效率。 Aiming at the problem of scarcity of on-board resources and low energy utilization efficiency in multi-beam satellite communication system, this paper proposed the power allocation algorithm in distributed satellite cluster net-work downlink considering both system power consumption and data rate. And the energy efficiency of the system is opti-mized by reasonable resource allocation.