用于行人轨迹预测的场景限制时空图卷积网络

目的 针对行人轨迹预测问题,已有的几种结合场景信息的方法基于合并操作通过神经网络隐式学习场景与行人运动的关联,无法直观地解释场景对单个行人运动的调节作用。除此之外,基于图注意力机制的时空图神经网络旨在学习全局模式下行人之间的社会交互,在人群拥挤场景下精度不佳。鉴于此,本文提出一种场景限制时空图卷积神经网络（scene-constrained spatial-temporal graph convolutional neural network,Scene-STGCNN）。方法 Scene-STGCNN由运动模块、基于场景的微调模块、时空卷积和时空外推卷积组成。运动模块以时空图卷积提取局部行人时空特征,避免了时空图神经网络在全局模式下学习交互的局限性。基于场景的微调模块将场景信息嵌入为掩模矩阵,用来调节运动模块生成的中间运动特征,具备实际场景下的物理解释性。通过最小化核密度估计下真实轨迹的负对数似然,增强Scene-STGCNN输出的多模态性,减少预测误差。结果 实验在公开数据集ETH （包含ETH和HOTEL）和UCY （包含UNIV、ZARA1和ZARA2）上与其他7种主流方法进行比较,就平均值而言,相对于性能第2的模型,平均位移误差（average displacement error,ADE）值减少了12%,最终位移误差（final displacement error,FDE）值减少了9%。在同样的数据集上进行了消融实验以验证基于场景的微调模块的有效性,结果表明基于场景的微调模块能有效建模场景对行人轨迹的调节作用,从而减小算法的预测误差。结论 本文提出的场景限制时空图卷积网络能有效融合场景和行人运动,在学习局部模式下行人交互的同时基于场景特征对轨迹特征做实时性调节,相比于其他主流方法,具有更优的性能。     

情景感知驱动的移动对象多模式轨迹预测技术综述

如何利用多源异构时空数据进行准确的轨迹预测并且反映移动对象的移动特性是轨迹预测领域的核心问题.现有的大多数轨迹预测方法是长序列轨迹模式预测模型,根据历史轨迹的特点进行预测,或将当前移动对象的轨迹位置放入时空语义场景根据历史移动对象轨迹预测位置.综述当前常用的轨迹预测模型和算法,涉及不同的研究领域.首先,阐述了多模式轨迹预测的主流工作,轨迹预测的基本模型类;其次,对不同类的预测模型进行总结,包括数学统计类、机器学习类、滤波算法,以及上述领域具有代表性的算法;再次,对情景感知技术进行了介绍,描述了不同领域的学者对情景感知的定义,阐述了情景感知技术所包含的关键技术点,诸如情景感知计算、情景获取和情景推理的不同类模型,分析了情景感知的不同分类、过滤、存储和融合以及它们的实现方法等.详细介绍了情景感知驱动的轨迹预测模型技术路线及各阶段任务的工作原理.给出了情景感知技术在真实场景中的应用,包括位置推荐,兴趣点推荐等,通过与传统算法对比,分析情景感知技术在此类应用中的优劣.详细介绍了情景感知结合LSTM (long short-term memory)技术应用于行人轨迹预测领域的新方法.最后,总结了...     

Effect of magnesium sulfate on the durability of limestone mortars based on quaternary blended cements

Currently, the use of blended cements incorporating various supplementary cementing materials, preserved in aggressive environments has become common. This paper describes the investigation results conducted on the evaluation of the resistance to magnesium sulfate solution (MgSO₄) of limestone mortars containing simultaneously; limestone filler, blast furnace slag and natural pozzolan. In this study, the deterioration of limestone mortars due to sulfate attack was evaluated by measuring changes in weight, length and compressive strength at the ages of 30, 60, 90, 120 and 180 days of immersion in exposure environments. The X-ray diffraction was also used in order to determine the different mineral phases. It is noteworthy that, the pH variation of the conservation solutions has been monitored during tests. The exposure solution was renewed monthly until the end of tests. The results showed that, the resistance to sulfate attack of mortars made with quaternary binders was better than that of mortars based on ordinary Portland cement.     

Leveraging Transfer Learning for Spatio-Temporal Human Activity Recognition from Video Sequences

Human Activity Recognition (HAR) is an active research area due to its applications in pervasive computing, human-computer interaction, artificial intelligence, health care, and social sciences. Moreover, dynamic environments and anthropometric differences between individuals make it harder to recognize actions. This study focused on human activity in video sequences acquired with an RGB camera because of its vast range of real-world applications. It uses two-stream ConvNet to extract spatial and temporal information and proposes a fine-tuned deep neural network. Moreover, the transfer learning paradigm is adopted to extract varied and fixed frames while reusing object identification information. Six state-of-the-art pre-trained models are exploited to find the best model for spatial feature extraction. For temporal sequence, this study uses dense optical flow following the two-stream ConvNet and Bidirectional Long Short Term Memory (BiLSTM) to capture long-term dependencies. Two state-of-the-art datasets, UCF101 and HMDB51, are used for evaluation purposes. In addition, seven state-of-the-art optimizers are used to fine-tune the proposed network parameters. Furthermore, this study utilizes an ensemble mechanism to aggregate spatial-temporal features using a four-stream Convolutional Neural Network (CNN), where two streams use RGB data. In contrast, the other uses optical flow images. Finally, the proposed ensemble approach using max hard voting outperforms state-of-the-art methods with 96.30% and 90.07% accuracies on the UCF101 and HMDB51 datasets.     

Fuzzy logic based assignable cause diagnosis using control chart patterns

Control chart patterns, besides determining the presence of assignable causes, also provide hints on the nature of assignable cause(s) present. Relating the patterns exhibited on the control chart to assignable causes is an ambiguous and vague task especially when multiple patterns co-exist. In this work, a rule based fuzzy inference system is developed for control chart to prioritize the control chart causes based on the accumulated evidence. When a process goes out of control, search for assignable causes can be assisted by the priorities assigned to the causes. For an in-control process, developing patterns can be tracked and preventive action can be taken to prevent the process from going out of control.     

Revisiting Zhang's 1D calibration algorithm

In recent years, the camera calibration using 1D patterns has been studied and improved by researchers all over the world. However, the progress in that area has been mainly in the sense of reducing the restrictions to the 1D pattern movement. On the other hand, the method's accuracy still demands improvements. In the present paper, the original technique proposed by Zhang is revisited and we demonstrate that the method's accuracy can be significantly improved, simply by analyzing and reformulating the problem. The numerical conditioning can be improved if a simple data normalization is performed. Furthermore, a non-linear solution based on the Partitioned Levenberg-Marquardt algorithm is proposed. That solution takes advantage of the problem's particular structure to reduce the computational complexity of the original method and to improve the accuracy. Tests using both synthetic and real images demonstrate that the calibration method using 1D patterns can be applied in practice, with accuracy comparable to other already traditional methods.     

LCMine: An efficient algorithm for mining discriminative regularities and its application in supervised classification

In this paper, we introduce an efficient algorithm for mining discriminative regularities on databases with mixed and incomplete data. Unlike previous methods, our algorithm does not apply an a priori discretization on numerical features; it extracts regularities from a set of diverse decision trees, induced with a special procedure. Experimental results show that a classifier based on the regularities obtained by our algorithm attains higher classification accuracy, using fewer discriminative regularities than those obtained by previous pattern-based classifiers. Additionally, we show that our classifier is competitive with traditional and state-of-the-art classifiers.     

The pragmatic quality of Resources‐ Events‐Agents diagrams: an experimental evaluation

The Resources‐Events‐Agents (REA) model is a semantic data model for the development of enterprise information systems. Although this model has been proposed as a benchmark for enterprise information modelling, only few studies have attempted to empirically validate the claimed benefits of REA modelling. Moreover, these studies focused on the evaluation of REA‐based system implementations rather than directly assessing the REA‐modelled conceptual schemas that these systems are based on. This paper presents a laboratory experiment that measured the user understanding of diagrammatic conceptual schemas developed using the REA model. The theoretical foundation for the hypotheses are cognitive theories that explain pattern recognition phenomena and the resulting reduction in cognitive effort for understanding conceptual schemas. The results of the experiment indicate a more accurate understanding of the business processes and policies modelled when users recognize the REA model’s core pattern of enterprise information in the diagram. The implication for modelling practice is that the use of the REA model improves the requirements engineering process by facilitating the user validation of conceptual schemas produced by analysts, and thus helps ensuring the quality of the enterprise information system that is developed or implemented.     

时空数据异常探测方法

以“k倍标准差”准则为基础,提出一种专题属性双重偏离的时空异常检测方法,在每个要素的空间邻近域里采用“k倍标准差”准则探测各时刻的空间异常数据,在每个空间异常数据的时间邻近域中,再次使用该准则判断该要素是否为时序异常,并将所有空间和时间邻近域上均表现为异常的数据定义为时空异常。实验结果表明,该方法是有效可行的。