内插螺旋立式上行管流场及传热性能的实验研究

通过粒子图像测速流场实验与传热实验相结合，研究了内插螺旋立式上行管的螺旋节距、丝径、中径比等结构参数在不同Re下对流场、阻力及传热性能的影响。结果表明，内插螺旋能够有效扰动和混合管内流体，使管内形成多个纵向旋涡的流体结构、增大管壁附近液体涡量，有利于强化传热。当Re相同时，管内平均流速v、Nu和综合换热性能PEC均随丝径增大而增大，随中径比减小而增大；随节距增大，3种参数均出现增大的趋势，节距大于20 mm后开始减小。管内流体的阻力f随丝径和节距增大而减小，随中径比增大而增大。综合比较，在较低Re时，节距p=20 mm、丝径e=1.6 mm、中径比D/d=0.75时综合传热效果最好。     

Study of the surfactant role in latex–aerogel systems by scanning transmission electron microscopy on aqueous suspensions

For insulation applications, boards thinner than 2 cm are under design with specific thermal conductivities lower than 15 mW m^?1 K^?1. This requires binding slightly hydrophobic aerogels which are highly nanoporous granular materials. To reach this step and ensure insulation board durability at the building scale, it is compulsory to design, characterise and analyse the microstructure at the nanoscale. It is indeed necessary to understand how the solid material is formed from a liquid suspension. This issue is addressed in this paper through wet‐STEM experiments carried out in an Environmental Scanning Electron Microscope (ESEM). Latex–surfactant binary blends and latex–surfactant–aerogel ternary systems are studied, with two different surfactants of very different chemical structures. Image analysis is used to distinguish the different components and get quantitative morphological parameters which describe the sample architecture. The evolution of such morphological parameters during water evaporation permits a good understanding of the role of the surfactant.     

基于循环一致性对抗网络的室内火焰图像场景迁移

基于深度学习的视频火灾探测模型的训练依赖于大量的正负样本数据,即火灾视频和带有干扰的场景视频。由于很多室内场合禁止点火,导致该场景下的火灾视频样本不足。本文基于生成对抗网络,将其他相似场景下录制的火焰迁移到指定场景,以此增广限制性场合下的火灾视频数据。文中提出将火焰内核预先植入场景使之具备完整的内容信息,再通过添加烟雾和地面反射等风格信息,完成场景与火焰的融合。该方法克服了现有多模态图像转换方法在图像转换过程中因丢失信息而造成的背景失真问题。同时为减少数据采集工作量,采用循环一致性生成对抗网络以解除训练图像必须严格匹配的限制。实验表明,与现有多模态图像转换相比,本文方法可以保证场景中火焰形态的多样性,迁移后的场景具有较高的视觉真实性,所得结果的FID与LPIPS值最小,分别为119.6和0.134 2。     

基于卷积神经网络的葡萄叶片氮含量识别方法

为实现对葡萄叶片氮素含量快速、便捷的识别，在卷积神经网络VGG-16网络结构基础上，将数据增广后的图像按不同梯度划分进行模型训练，通过十折交叉验证法探究最佳的训练集与验证集分配比例，并构建4个不同深度的网络模型进行训练对比，采用全局平均池化代替全连接层约简网络参数量。训练结果表明，氮含量梯度设为0.70%、0.35%和0.175%时，室内简单背景识别准确率分别为85.9%、76.2%和71.1%；晴天室外复杂背景下识别准确率分别为44.6%、35.0%和30.4%。研究结果表明利用VGG-16建立的网络学习模型对葡萄叶片氮含量识别提供了一种新的便捷方法，对农业信息化和智能化技术应用具有一定促进作用。     

R2D2: A scalable deep learning toolkit for medical imaging segmentation

Deep learning has gained a significant popularity in recent years thanks to its tremendous success across a wide range of relevant fields of applications, including medical image analysis domain in particular. Although convolutional neural networks (CNNs) based medical applications have been providing powerful solutions and revolutionizing medicine, efficiently training of CNNs models is a tedious and challenging task. It is a computationally intensive process taking long time and rare system resources, which represents a significant hindrance to scientific research progress. In order to address this challenge, we propose in this article, R2D2, a scalable intuitive deep learning toolkit for medical imaging semantic segmentation. To the best of our knowledge, the present work is the first that aims to tackle this issue by offering a novel distributed versions of two well-known and widely used CNN segmentation architectures [ie, fully convolutional network (FCN) and U-Net]. We introduce the design and the core building blocks of R2D2. We further present and analyze its experimental evaluation results on two different concrete medical imaging segmentation use cases. R2D2 achieves up to 17.5× and 10.4× speedup than single-node based training of U-Net and FCN, respectively, with a negligible, though still unexpected segmentation accuracy loss. R2D2 offers not only an empirical evidence and investigates in-depth the latest published works but also it facilitates and significantly reduces the effort required by researchers to quickly prototype and easily discover cutting-edge CNN configurations and architectures.     

基于视觉颜色感知——光学相似的图像去雾方法

针对雾霾环境影响致使户外获取的图像质量严重下降问题，提出了一种基于视觉颜色感知—光学相似的图像去雾方法。充分利用人眼感知颜色的视觉机理，结合图像的相似性原理，构造了光学相似度函数，建立了新的基于视觉颜色感知—光学相似的图像去雾模型并设计相关算法，进而进行仿真验证。仿真实验结果表明，提出的方法在对有雾图像清晰化处理过程中效果明显，并与现有的图像去雾方法在主观视觉和客观量化方面进行图像去雾效果对比，进一步表明提出的方法在清晰化含雾图像处理中取得了较好的效果。     

伪标签置信选择的半监督集成学习视频语义检测

在视频语义检测中，有标记样本不足会严重影响检测的性能，而且伪标签样本中的噪声也会导致集成学习基分类器性能提升不足。为此，提出一种伪标签置信选择的半监督集成学习算法。首先，在三个不同的特征空间上训练出三个基分类器，得到基分类器的标签矢量；然后，引入加权融合样本所属某个类别的最大概率与次大概率的误差和样本所属某个类别的最大概率与样本所属其他各类别的平均概率的误差，作为基分类器的标签置信度，并融合标签矢量和标签置信度得到样本的伪标签和集成置信度；接着，选择集成置信度高的样本加入到有标签的样本集，迭代训练基分类器；最后，采用训练好的基分类器集成协作检测视频语义概念。该算法在实验数据集UCF11上的平均准确率到达了83.48%，与Co-KNN-SVM算法相比，平均准确率提高了3.48个百分点。该算法选择的伪标签能体现样本所属类别与其他类别的总体差异性，又能体现所属类别的唯一性，可减少利用伪标签样本的风险，有效提高视频语义概念检测的准确率。