ID 原文 译文
24805 同时，利用上采样和过滤技术扩展了最佳高斯整数序列的长度和自由度。 The length and degree of freedom of the perfect Gaussian integer sequences are extended by up-sampling and filtering.
24806 本文方法能得到大量适于高速通信系统的最佳高斯整数序列，扩展了通信地址码的选择范围。 A large number of perfect Gaussian integer sequences obtained in this paper are suitable for high speed communication applications, which expands the selection range of address codes.
24807 为了有效降低室内基于广义空移键控(Generalized Space Shift Keying，GSSK)调制技术的可见光通信(Visible Light Communications，VLC)系统 LED选择算法的复杂度，提高系统的 LED选择速度，提出了一种基于支持向量机(Support Vector Machine，SVM)的机器学习低复杂度高效率 LED选择算法。 In order to reduce the complexity of LED selection algorithm in generalized space shift keying (GSSK) aided indoor visible light communication (VLC) system, a support vector machine (SVM) assisted low complexity and high efficiency machine learning LED selection algorithm is proposed for the considered GSSK-VLC system.
24808 通过将室内 GSSK⁃VLC系统的 LED选择等价建模为多分类的机器学习问题，利用核 SVM 构建 LED选择的最优化问题，通过对偶理论，获得原问题的二次凸规划对偶问题，从而高效的获取 SVM 的最优分类参数 。最后，通过学习训练获得的最优分类参数实现对任意给定用户信道信息的在线天线选择。 By modeling the LED selection in indoor GSSK-VLC system as a multi-classification problem, an optimization problem is constructed by utilizing kernel SVM. After the optimal parameters of the learning system are obtained, the LED selection procedure can be accomplished efficiently for any given user's channel state information.
24809 通过计算机仿真和复杂度分析，与传统的 LED 选择算法相比，本文提出的算法能够在实现在线LED选择的同时保持低误码率(Bit Error Ratio，BER)性能，表明了该算法的有效性。 Simulation results and complexity analysis show that, compared with traditional LED selection algorithms, the proposed SVM aided LED selection algorithm can achieve an ideal bit error ratio (BER) performance while having considerably lower complexity.
24810 传统方法在视觉策略网络中只关注实体，不能够推理出实体和属性之间的联系，在语言策略网络存在暴露偏差和误差累计问题。 Traditional methods only focus on entities in the visual strategy network and cannot deduce the relationship between entities and attributes. There are problems of exposure bias and error accumulation in the language strategy network.
24811 为此，提出了一个基于强化学习的多层级视觉融合网络模型。 Therefore, this paper proposes a multilevel visual fusion network model based on reinforcement learning.
24812 在视觉策略网络中通过多层级神经网络模块将视觉特征转化为视觉知识的特征集。 In the visual strategy network, multilevel subneural network module is used to transform visual features into feature sets of visual knowledge.
24813 融合网络生成使描述语句更加流畅的虚词，用于视觉策略网络和语言策略网络的互动。 The fusion network generates the function words which make the description sentences more fluent and can be used for the interaction between the visual strategy network and the language strategy network.
24814 在语言策略网络中使用基于强化学习的自批评策略梯度算法对视觉融合网络实现端到端的优化。 The gradient algorithm of self-criticism strategy based on reinforcement learning is used to optimize the visual fusion network end-to-end.