ID	原文	译文
19245	针对复杂声学环境下，现有目标声源定位算法精度低的问题，该文提出了一种基于时频单元选择的双耳目标声源定位算法。	The performance of the existing target localization algorithms is not ideal in complex acousticenvironment. In order to improve this problem, a novel target binaural sound localization algorithm ispresented.
19246	该算法首先利用双耳目标声源的频谱特征训练1个基于深度学习的时频单元选择模型，然后使用时频单元选择器从双耳输入信号中提取可靠的时频单元，减少非目标时频单元对定位精度的负面影响。	First, the algorithm uses binaural spectral features as input of a time-frequency units selector basedon deep learning. Then, to reduce the negative impact of the time-frequency unit belonging to noise on the localization accuracy, the selector is employed to select the reliable time-frequency units from binaural input sound signal.
19247	同时，基于深度神经网络的定位系统将双耳空间线索映射到方位角的后验概率。	At the same time, a Deep Neural Network (DNN)-based localization system maps the binauralcues of each time-frequency unit to the azimuth posterior probability.
19248	最后，依据与可靠时频单元相对应的后验概率完成目标语音的声源定位。	Finally, the target localization iscompleted according to the azimuth posterior probability belonging to the reliable time-frequency units.
19249	实验结果表明，该算法在低信噪比和各种混响环境，特别是存在与目标声源类似的噪声环境下目标声源的定位精度得到明显改善，性能优于对比算法。	Experimental results show that the performance of the proposed algorithm is better than comparison algorithmsand achieves a significant improvement in target localization accuracy in low Signal-to-Noise Ratio(SNR) andvarious reverberation environments, especially when there is noise similar to the target sound source.
19250	该文提出一种基于头脑风暴智能优化算法的BP神经网络模糊图像复原方法(OBSO-BP)。	A kind of restoration method of BP neural network fuzzy image based on Optimized Brain Storming intelligent Optimized(OBSO-BP) algorithm is proposed in this paper.
19251	该方法在聚类和变异两方面优化了头脑风暴智能算法，利用头脑风暴优化算法易于解决多峰高维函数问题的特点，自动搜寻BP神经网络更佳的初始权值和阈值，以减少BP网络对其初始权值和阈值的敏感性，避免网络陷入局部最优解，增加网络的收敛速度，减小网络误差，提高图像还原质量。	With the method of brain stormingintelligent optimized algorithm which is optimized in both clustering and variation, issues of multi-peak high-dimensional function is easily solved. This method optimizes brain storming intelligence algorithm from twoaspects of clustering and mutation. This method makes use of the characteristics of brain storming optimizationalgorithm, which is easy to solve multi-peak and high-dimensional function problems, to automatically searchfor better initial weights and thresholds of BP neural network, thus reducing the sensitivity of BP network toits initial weights and thresholds, avoiding the network falling into local optimal solution, increasing theconvergence speed of the network and reducing the network error and improving the quality of imagerestoration.
19252	该文采用20张不同的图像，对其模糊图像分别进行维纳滤波复原(Wiener)、基于头脑风暴算法的维纳滤波复原(Wiener-BSO)、BP神经网络复原以及基于头脑风暴算法的BP神经网络(BSO-BP)图像复原实验。	Twenty different images are adopted to the image restoration experiment of their fuzzy images withWiener filtering restoration(Wiener), Wiener filtering restoration based on optimized Brain Storming intelligentOptimized algorithm(Wiener-BSO), BP neural network restoration and BP neural network restoration based onoptimized Brain Storming intelligent Optimized algorithm(BSO-BP).
19253	实验结果表明，该方法能够取得更好的图像复原效果。	Results show that a better effect of imagerestoration can be achieved with this method.
19254	双基雷达具有隐蔽性高、抗干扰性能强等优点，在现代电子战中发挥重要作用。	Bistatic radar has the advantages of high concealment and strong anti-interference performance, and plays an important role in modern electronic warfare.