ID 原文 译文
39376 然后，分析循环频率与四阶累积量的代数关系; Then, the algebraic relationship between the cyclic frequency and the fourth-order cumulants is analyzed.
39377 最后，得出四阶累积量的峰值受原型滤波器影响且仅在其循环频率等于子载波频率处存在。通过检测四阶累积量峰值的位置和个数能有效估计子载波频率和数目。 Finally, because the value of the fourth-order cumulants' peak is affected by the prototype filter and only exists when the cyclic frequency equals to the subcarrier frequency, the subcarrier frequency and number of FBMC signals can be estimated by detecting the locations and number of fourth-order cumulants' peaks.
39378 仿真表明，当信噪比大于2 dB时，算法对FBMC-OQAM信号的子载波参数盲估计正确概率为98%。 Simulation results show that when the SNR is greater than 2 dB, the correct probability of blind estimation of the subcarrier parameters of the FBMC-OQAM signal is 98%.
39379 在基于深度卷积神经网络的目标检测方法中，模型的参数量动辄数十兆字节，在计算资源有限的移动终端等边缘设备中部署这样的大模型比较困难。 In the object detection method based on deep convolutional neural networks, the parameter amount of its model is often tens of megabytes, so it is difficult to deploy such a large model in edge devices such as mobile terminals with limited computing resources.
39380 为了解决这个问题，本文在Single Shot MultiBox Detector(SSD)的基础上联合轻量化网络设计和参数量化两种技术来实现网络模型的轻量化。 In order to solve this problem, this paper combines lightweight network design and parameter quantification to achieve the lightweight of the network on the basis of one-stage object detector(SSD).
39381 首先，基于ResNet50和MobileNet我们重新设计了SSD目标检测框架，并训练了一个全精度参数模型。 First, we redesigned the SSD object detection framework based on ResNet50 and MobileNet and trained a model with full precision weights.
39382 然后，在全精度参数模型的基础上，采取逐块量化的策略将特征提取层中卷积层的参数精度降低到三值(零和正负一)。 Then, based on the full-precision weight model, a block-by-block quantization strategy is adopted to reduce the weight accuracy of the convolutional layer in the feature extraction layer to three values(zero, plus and minus one).
39383 实验结果表明，本文提出的联合方案在Pascal VOC2007数据集上测试能够达到72.54%的mAP，和其他业界领先的轻量级目标检测方法相比检测精度更高且能使模型占用的内存空间更小。 The experimental results show that the joint scheme proposed in this paper can test 72.54% mAP on the Pascal VOC2007 dataset. Compared with other industry-leading lightweight object detection methods, the detection accuracy is higher and the model can occupy less memory space.
39384 基于信道状态信息的室内轨迹跟踪技术近几年来备受关注。 Indoor tracking technology based on Channel State Information received much attention in recent years.
39385 为了利用现有的商用WiFi设备实现实时的轨迹跟踪，本文采用了一发一收的室内轨迹跟踪系统，主要利用通过信道状态信息计算得到的多维参数包括飞行时间、到达角度、多普勒频移和信道衰减来计算得到相应的位置信息， In order to realize real-time trajectory tracking by using existing commercial WiFi devices, this paper adopted indoor tracking system with one-transmitter and one-receiver, the multidimensional parameters calculated by Channel State information, including Time of Flight, Arrival of Angle, Doppler Frequency Shift and channel attenuation which were used to calculate the corresponding position information,