ID 原文 译文
57708 通过多周期迭代分析碰撞概率，建立了理论模型，有效地解决了大规模传感器节点并发广播信道碰撞严重的问题. The collision probability was analyzed by multi-period iteration，and a theoretical model was established which effectively solves the serious channel collision in concurrent broadcasting oflarge-scale sensor nodes.
57709 使用蒙特卡洛法仿真评估延迟、容量指标，分析了影响性能的关键参数，模型计算结果与仿真结果一致. Monte Carlo simulation was created to evaluate indicators such as latency，capacity，and power consumption. The key parameters affecting the performance were analyzed.
57710 与无应答式的广播协议相比，所采用的广播应答机制可以高效确认，降低了 23% 的网络延迟和 90% 的广播碰撞，并使系统容量提升 1 倍. The modelcalculation results are consistent with the simulation results. Compared with the unresponsive broadcastingprotocol，the broadcast response mechanism can effectively confirm，reduce 23% network delay and 90%broadcast collision，and increase the system capacity by 100% .
57711 针对复杂场景下目标容易丢失的问题，提出了一种基于深度残差网络( ＲesNet) 特征的尺度自适应视觉目标跟踪算法.首先，通过 ＲesNet 提取图像感兴趣区域的多层深度特征，考虑到修正线性单元( ＲeLU) 激活函数对目标特征的抑制作用，在 ＲeLU 函数之前选取用于提取目标特征的卷积层; Because the objects are easy to be lost in complex scenes，a scale adaptive visual object track- ing algorithm based on deep residual network ( ＲesNet) features is proposed. Firstly，the ＲesNet is used to extract the multi-layer deep features of the image region of interest. Considering the restraining effect of rectified linear units ( ＲeLU) activation function on target features，only the convolutional layers before ＲeLU function are selected.
57712 然后，在提取的多层特征上分别构建基于核相关滤波的位置滤波器，并对得到的多个响应图进行加权融合，选取响应值最大的点即为目标中心位置. Secondly，the translation filters based on kernelized correlation filter are constructed in the extracted multi-layer features，and then the weighted fusion of the multiple response maps is carried out to obtain the target position with the largest response value.
57713 目标位置确定后，对目标进行多个尺度采样，分别提取不同尺度图像的方向梯度直方图( fHOG) 特征，在此基础上构建尺度相关滤波器，从而实现对目标尺度的准确估计. After the target location is determined，the target is sampled at multiple scales，and the felzenszwalb histogram of oriented gradients ( fHOG) features of different scale images are extracted separately. On this basis，a scale correlation fil- ter is constructed to estimate the target scale accurately.
57714 在视频集 OTB100 中与其他 6 种相关算法进行了比较，实验结果表明，所提算法取得了较高的跟踪成功率和精确度，能够较好地适应目标的尺度变化、背景干扰等复杂场景. Comparing with six related algorithms in OTB100，an experiment is carried. It is shown that the proposed algorithm achieves high tracking success rate and accuracy，and can adapt to scale variation，background clutter and other complex scenes.
57715 提出了一种鲁棒网络流量分类及新类型的发现算法.网络流量一般为高维数据，且在网络流量收集过程中易受到网络波动或网络攻击的影响，为此，在堆栈自编码器的基础上，基于互相关熵理论提出了一种新的网络模型进行数据的特征提取，通过基于阈值的主动学习分类算法进行分类，达到识别新应用类型的目的. robust network traffic classification and new type discovery algorithm is proposed by this pa- per，which is based on sparse autoencoder to extract feature features and classify based on threshold- based active learning classification algorithm. In addition，to achieve the purpose of identifying new ap- plication types，the excellent performance of the proposed algorithm through comparative experiments is verified.
57716 对比实验结果表明，所提算法中分类算法的准确度可达到 91. Among them，the accuracy of the classification algorithm can reach 91.
57717 08% ，对新应用类型的识别度可达到 98. 08% ; the recognition of new application types can reach 98.