ID 原文 译文
19345 仿真结果表明该算法能实现对用户终端的功率有效控制，提升系统能效。 The simulation results show that the algorithm can effectively control the power of the user terminal and improve system energy efficient.
19346 人类面部表情是其心理情绪变化的最直观刻画，不同人的面部表情具有很大差异，现有表情识别方法均利用面部统计特征区分不同表情，其缺乏对于人脸细节信息的深度挖掘。 Facial expression is the most intuitive description of changes in psychological emotions, and different people have great differences in facial expressions. The existing facial expression recognition methods use facialstatistical features to distinguish among different expressions, but these methods are short of deep explorationfor facial detail information.
19347 根据心理学家对面部行为编码的定义可以看出，人脸的局部细节信息决定了其表情意义。 According to the definition of facial behavior coding by psychologists, it can beseen that the local detail information of the face determines the meaning of facial expression.
19348 因此该文提出一种基于多尺度细节增强的面部表情识别方法，针对面部表情受图像细节影响较大的特点，提出利用高斯金字塔提取图像细节信息，并对图像进行细节增强，从而强化人脸表情信息。 Therefore, a facial expression recognition method based on multi-scale detail enhancement is proposed, because facial expression is much more affected by the image details than other information, the method proposed in this paper extracts the image detail information with the Gaussian pyramid firstly, thus the image is enhanced in detail to enrich the facial expression information.
19349 针对面部表情的局部性特点，提出利用层次结构的局部梯度特征计算方法，描述面部特征点局部形状特征。 Secondly, for the local characteristics of facial expressions, a local gradient feature calculation method is proposed based on hierarchical structure to describe the local shape features of facial feature points.
19350 最后，使用支持向量机(SVM)对面部表情进行分类。 Finally, facial expressions are classified using a Support Vector Machine (SVM).
19351 该文在CK+表情数据库中的实验结果表明，该方法不仅验证了图像细节对面部表情识别过程的重要作用，而且在小规模训练数据下也能够得到非常好的识别结果，表情平均识别率达到98.19%。 Theexperimental results in the CK+ expression database show that the method not only proves the important roleof image detail in facial expression recognition, but also obtains very good recognition results under small-scaletraining data. The average recognition rate of expressions reaches 98.19%.
19352 目前大多数迁移学习方法在利用源域数据辅助目标域数据建模时，通常假设源域中的数据均与目标域数据相关。 Most current transfer learning methods are modeled by utilizing the source data with the assumptionthat all data in the source domain are equally related to the target domain.
19353 然而在实际应用中，源域中的数据并非都与目标域数据的相关程度一致，若基于上述假设往往会导致负迁移效应。 In many practical applications,however, this assumption may induce negative learning effect when it becomes invalid.
19354 为此，该文首先提出分类误差一致性准则(CCR)，对源域与目标域分类误差的概率分布积分平方误差进行最小化度量。 To tackle this issue, byminimizing the integrated squared error of the probability distribution of the source and target domainclassification errors, the Classification-error Consensus Regularization (CCR) is proposed.