基于人体关节点数据的攻击性行为识别

针对人体攻击性行为识别问题，提出一种基于人体关节点数据的攻击性行为识别方法。首先，利用OpenPose获得单帧图像中的人体关节点数据，并通过最近邻帧特征加权法和分段多项式回归完成由人体自遮挡和环境因素所导致缺失值的补全；然后，对每个人体定义动态"安全距离"阈值，如果两人真实距离小于阈值，则构建行为特征矢量，其中包括帧间人体重心位移、人体关节旋转角角速度和发生交互时的最小攻击距离等；最后，提出改进的LightGBM算法w-LightGBM，并对攻击性行为进行识别。采用公共数据集UT-interaction对所提出的攻击性行为分类识别方法进行测试实验，准确率达到95.45%。实验结果表明，所提方法能够有效识别各种角度的攻击性行为。     

Emotion recognition at the edge with AI specific low power architectures

Nowadays Deep Learning is applied in almost every research field and helps getting amazing results in a great number of challenging tasks. The main problem is that this kind of learning and consequently Neural Networks that can be defined deep, are resource intensive. They need specialized hardware to perform computation in a reasonable time. Many tasks are mandatory to be as much real-time as possible . It is needed to optimize many components such as code, algorithms, numeric accuracy and hardware, to make them “efficient and usable”. All these optimizations can help us to produce incredibly accurate and fast learning models. The paper reports a study in this direction for the challenging face detection and emotion recognition tasks.     

Flow regime recognition in a long pipeline-riser system based on signals at the top of the riser

To identify conveniently multiphase flow regimes in subsea pipeline-risers, we study in this paper experimentally two-phase flows in a 1657 m long pipeline with an S-shaped riser to simulate field experiment, within a wide range of gas and liquid velocities. Three flow regimes, namely severe slugging, transitional flows, and stable flows, are analyzed based on three differential pressure and one pressure signals at the top of the riser; comparatively speaking, the positions of these signals in the experimental system are similar to those of the sea level signals in industrial fields, which are easy and less expensive to obtain. The obtained signals are decomposed into six scales via a multi-scale wavelet analysis, and further four statistical parameters on each scale are extracted, including mean values, standard deviations, ranges, and mean values of absolute. We compared the effects of six SVM classifiers with different kernel functions on the recognition rate of flow regimes, and it is found the recognition rates of SVM classifier with quadratic and cubic kernel functions are the highest. Further, the principal component analysis is employed to reduce the dimension of statistical parameters and it indicates that the recognition rate tends to increase with the rising number of principal components from 1 to 6, and it remains constant if the principal component number is further increased. Moreover, The results suggest that the recognition rate obtained from the pressure difference between the top of the riser and the separator peaks, and then it comes that from the pressure signal at the top of the riser, and that for the pressure difference signal at the top of the riser is the least satisfying one. As for the optimal differential pressure signals between the top of the riser and the separator, the results show that the recognition rate increases rapidly from 70.2% to 90.4% when the sample duration rising from 2.3 s to 18.6 s, and when the sample duration exceeds 74.4 s, the recognition rate exceeds 92.9% and remains unchanged.     

基于多尺度双线性卷积神经网络的多角度下车型精细识别

针对多角度下车辆出现一定的尺度变化和形变导致很难被准确识别的问题，提出基于多尺度双线性卷积神经网络（MS-B-CNN）的车型精细识别模型。首先，对双线性卷积神经网络（B-CNN）算法进行改进，提出MS-B-CNN算法对不同卷积层的特征进行了多尺度融合，以提高特征表达能力；此外，还采用基于中心损失函数与Softmax损失函数联合学习的策略，在Softmax损失函数基础上分别对训练集每个类别在特征空间维护一个类中心，在训练过程中新增加样本时，网络会约束样本的分类中心距离，以提高多角度情况下的车型识别的能力。实验结果显示，该车型识别模型在CompCars数据集上的正确率达到了93.63%，验证了模型在多角度情况下的准确性和鲁棒性。     

加权信息熵与增强局部二值模式结合的人脸识别

针对人脸识别因光照、姿态、表情、遮挡及噪声等多种因素的影响而导致的识别率不高的问题，提出一种加权信息熵（IEw）与自适应阈值环形局部二值模式（ATRLBP）算子相结合的人脸识别方法（IE （w） ATR-LBP）。首先，从原始人脸图像分块提取信息熵，得到每个子块的IEw；然后，利用ATRLBP算子分别对每个人脸子块提取特征从而得到概率直方图；最后，将各个块的IEw与概率直方图相乘，再串联成为原始人脸图像最后的特征直方图，并利用支持向量机（SVM）对人脸进行识别。在AR人脸库的表情、光照、遮挡A和遮挡B四个数据集上，IE （w） ATR-LBP方法分别取得了98.37%、94.17%、98.20%和99.34%的识别率。在ORL人脸库上，IE （w） ATR-LBP方法的最大识别率为99.85%；而且在ORL人脸库5次不同训练样本的实验中，与无噪声时相比，加入高斯和椒盐噪声后的平均识别率分别下降了14.04和2.95个百分点。实验结果表明，IE （w） ATR-LBP方法能够有效提高人脸在受光照、姿态、遮挡等影响时的识别率，尤其是存在表情变化及脉冲类噪声干扰时的识别率。     

Computational personality recognition from Facebook text: psycholinguistic features,words and facets

ABSTRACT

Advances in the Natural Language Processing (NLP) and machine learning fields have led to the development of automated methods for the recognition of personality traits from text available from social media and similar sources. Systems of this kind exploit the close relation between lexical knowledge and personality models – such as the well-known Big Five model – to provide information about the author of an input text in a non-intrusive fashion, and at a low cost. Although now a well-established research topic in the field, the computational recognition of personality traits from text still leaves a number of research questions worth further exploration. In particular, this paper attempts to shed light on three main issues: (i) whether we may develop psycholinguistics-motivated models of personality recognition when such knowledge sources are not available for the target language under consideration; (ii) whether the use of psycholinguistic knowledge may be still superior to contemporary word vector representations; and (iii) whether we may infer certain personality facets from a corpus that does not explicitly convey this information. In this paper these issues are dealt with in a series of individual experiments of personality recognition from Facebook text, whose initial results should aid the future development of more robust systems of this kind.     

Facial expression recognition of intercepted video sequences based on feature point movement trend and feature block texture variation

Facial Expression Recognition (FER) is an important subject of human–computer interaction and has long been a research area of great interest. Accurate Facial Expression Sequence Interception (FESI) and discriminative expression feature extraction are two enormous challenges for the video-based FER. This paper proposes a framework of FER for the intercepted video sequences by using feature point movement trend and feature block texture variation. Firstly, the feature points are marked by Active Appearance Model (AAM) and the most representative 24 of them are selected. Secondly, facial expression sequence is intercepted from the face video by determining two key frames whose emotional intensities are minimum and maximum, respectively. Thirdly, the trend curve which represents the Euclidean distance variations between any two selected feature points is fitted, and the slopes of specific points on the trend curve are calculated. Finally, combining Slope Set which is composed by the calculated slopes with the proposed Feature Block Texture Difference (FBTD) which refers to the texture variation of facial patch, the final expressional feature are formed and inputted to One-dimensional Convolution Neural Network (1DCNN) for FER. Five experiments are conducted in this research, and three average FER rates 95.2%, 96.5%, and 97% for Beihang University (BHU) facial expression database, MMI facial expression database, and the combination of two databases, respectively, have shown the significant advantages of the proposed method over the existing ones.     

基于堆叠深度卷积沙漏网络的步态识别

提出了一种基于堆叠深度卷积沙漏网络的步态识别方法。为了解决人体建模中关节点准确定位的问题，采用基于深度卷积的沙漏网络来提取步态图上的关节点坐标，并计算肘关节与膝关节的角度作为运动特征。为了解决行走速度变化带来的影响，采用动态时间规整（Dynamic Time Warping）对特征序列进行距离计算。通过最近邻分类器对结果进行准确分类。该方法在公共CASIA-B数据集与TUM-GAID数据集上进行了验证并与其他方法进行比较，结果表明该方法有较高的识别率。     

基于CNN的人体姿态识别

姿态识别是人机交互中重要的研究课题之一，随着机器学习与神经网络的发展，研究的方式和成果趋于多样化，姿态识别的应用价值也日趋广泛。本文通过构建卷积神经网络模型，该模型共有11层，在对采样的数据集中5种人体姿态进行卷积与池化操作，最后进入全连接层进行分类，从而完成对数据集的训练和识别。结果显示，相较于机器学习方法，该模型的识别性能更加优秀，且免去了复杂的特征提取方式设计，让网络自身提取特征进行识别分类，效果更好。