ID 原文 译文
56458 实验结果表明,通过约束标量场的变化范围,能在保持其方向不变的情况下,有效地控制去旋程度; The results indicate that the proposed method can reduce the curl of the frame fields efficiently while keeping their directions as well as controlling the extent of their length.
56459 将该标架场运用于已有重网格化方法,可得到方向和密度可控的六面体主导的混合网格. By applying such frame fields to existing hexahedral-dominantremeshing methods, direction- and size-controllable hexahedral-dominant meshes can be obtained.
56460 此外,该方法只需求解一个带界约束的二次凸规划问题,鲁棒性强且易于计算. Besides, theproposed method deduces to a convex quadratic programming problem with simple box constraints, which iseasy-to-solve
56461 传统基于分类学习的监督哈希方法并不能完全满足哈希检索技术需求,但是线性判别分析却能够在一定程度上做到这一点. The conventional supervised hashing methods based on classification do not entirely meet the requirements of the hashing technique; on the other hand, the linear discriminant analysis (LDA) approach doessatisfy such demands.
56462 本文提出将线性判别分析作为深度网络的优化目标,以端到端训练的方式学习有效的哈希编码. In this paper, we propose to perform the LDA objective over deep networks to learn effi?cient hashing codes in a truly end-to-end fashion.
56463 但是,直接以上述目标训练神经网络就必须解决具有较高计算复杂度的特征值分解问题. However, a complicated eigenvalue decomposition within eachmini-batch in every epoch has to be faced with when simply optimizing the deep network with respect to the LDAobjective.
56464 在本文中,线性判别分析目标被转化为一个简单的最小均方问题,这种转化可以解决上述问题,同时可以利用成熟的优化方法优化网络. Here, the LDA objective is transformed into a simple least-square problem, which naturally overcomesthe intractable problems and can be easily solved by an off-the-shelf optimizer.
56465 这种基于线性判别分析的深度网络拓展可以弥补传统判别分析在简单线性投影和特征学习上的劣势. Such deep extension can alsoovercome the weaknesses of LDA hashing involving the limited linear projection and feature learning.
56466 本文在3个基准数据集上进行大量对比实验,相对于传统线性判别分析,本文所提方法在检索基准指标上有70%的提升,并超过大多数基于深度模型的哈希方法,这些实验结果证明了本文方法的有效性. Numerousexperiments were conducted on three benchmark datasets. The proposed deep LDA hashing approach exhibitsan improvement of nearly 70 points for the CIFAR-10 dataset over the conventional strategy. Additionally, theproposed approach is found to be superior to several state-of-the-art methods for various metrics.
56467 雷达成像是获取目标精细结构特征的重要技术途径. Radar imaging is an essential technique to acquire fine structural characteristics of a target.