ID 原文 译文
41006 针对这种特征，设计了一种面向存储的高性能双向SPI传输接口，并定义了一套可扩展的存储传输协议，实现命令级别和数据级别的传输操作。 In response to this feature, a high-performance storage-oriented bidirectional SPI transmission interface is designed, and a set of extensible storage transmission protocols is defined to realize the transmission operations at the command level and the data level.
41007 该设计基于FPGA实现，比对仿真结果表明:该传输机制支持从设备主动传输，实现真正双向SPI通信，其传输效率相对当前主流Flash存储设备具有明显优势。 The design is based on FPGA implementation.The comparison simulation results show that the transmission mechanism supports active transmission from the device to achieve true two-way SPI communication, and its transmission efficiency has obvious advantages over mainstream flash memory storage devices.
41008 为解决在正交幅度调制-正交频分复用调制(QAM-OFDM)系统中的自适应比特分配算法复杂度过高的问题，提出了一种子载波比特分配方案。 In order to solve the problem of excessive complexity of the adaptive bit allocation algorithm in the orthogonal amplitude modulation-orthogonal frequency division multiplexing modulation(QAM-OFDM) system, an adaptive bit allocation algorithm is proposed.
41009 基于边缘自适应准则，通过对信道增益的预处理，在迭代分配中利用与运算以及移位操作来减少贪婪(greedy)算法中的计算和搜寻操作。 Based on the edge adaptation criterion, with the preprocessing of the channel gain, the iterative allocation uses AND operations and shift operations to reduce the calculation and sorting operations in the greedy algorithm.
41010 仿真结果显示:在相同的仿真环境下，改进算法与贪婪算法有着几乎相同的比特分配效果以及误比特率性能，并且较贪婪算法所用的时间更少，由此验证了改进算法具有很高的实用性。 In the same simulation environment, the simulation results show that the improved algorithm has almost the same bit allocation effect as the greedy algorithm, and it takes less time than the greedy algorithm, thus verifying that the improved algorithm has high practicability.
41011 在跨语料库语音情感识别中，训练和测试数据分布的差异变得非常明显，导致验证和测试性能差别很大. The difference in data distributions becomes very clear when the training and testing data come from different corpora, causing a large performance gap between validation and testing performance.
41012 针对该问题，提出一种基于对抗训练的跨语料库语音情感识别方法. To solve this problem, a cross-corpus speech emotion recognition method based on adversarial training is proposed.
41013 该方法通过语料库之间的对抗训练能有效地缩小不同语料库之间的差异，提升模型对域不变情感特征的提取能力. The proposed method can effectively eliminate the differences between different corpora with the adversarial training of corpora, and improve the extracting ability of domain-invariant emotion features.
41014 同时，通过引入多头自注意力机制，对语音序列中不同位置元素之间的依赖关系进行序列建模，增强序列中情感显著特征的提取能力. At the same time, model the relative dependence of different position elements in the speech sequence to enhance the emotion-salient features extracting ability of the sequence by introducing the multi-head attention mechanism.
41015 在以IEMOCAP为源域、MSP-IMPRO为目标域和在以MSP-IMPRO为源域、IEMOCAP为目标域上的实验表明，所提出方法的相对UAR性能相比于基准方法分别提升了0.91%～12.22%和2.27%～6.90%. When the experiment applies IEMOCAP as the source domain and MSP-IMPRO as the target domain, the results are superior to the benchmark methods about 0.91%～12.22%.Meanwhile, the experiment applies MSP-IMPRO as the source domain and IEMOCAP as the target domain, the results also achieve better performance than the benchmark methods about 2.27%～6.90%.