An on-line,real-time learning method for detecting anomalies in videos using spatio-temporal compositions

This paper presents an approach for detecting suspicious events in videos by using only the video itself as the training samples for valid behaviors. These salient events are obtained in real-time by detecting anomalous spatio-temporal regions in a densely sampled video. The method codes a video as a compact set of spatio-temporal volumes, while considering the uncertainty in the codebook construction. The spatio-temporal compositions of video volumes are modeled using a probabilistic framework, which calculates their likelihood of being normal in the video. This approach can be considered as an extension of the Bag of Video words (BOV) approaches, which represent a video as an order-less distribution of video volumes. The proposed method imposes spatial and temporal constraints on the video volumes so that an inference mechanism can estimate the probability density functions of their arrangements. Anomalous events are assumed to be video arrangements with very low frequency of occurrence. The algorithm is very fast and does not employ background subtraction, motion estimation or tracking. It is also robust to spatial and temporal scale changes, as well as some deformations. Experiments were performed on four video datasets of abnormal activities in both crowded and non-crowded scenes and under difficult illumination conditions. The proposed method outperformed all other approaches based on BOV that do not account for contextual information.     

动作识别中局部时空特征的运动表示方法研究

近年来,基于局部时空特征的运动表征方法已被越来越多地运用于视频中的动作识别问题,相关研究人员已经提出了多种特征检测和描述方法,并取得了良好的效果。但上述方法在适应摄像头移动、光照以及穿着变化等方面还存在明显不足。为此,提出了基于时空兴趣点局部时空特征的运动表示方法,实现了基于时空单词的动作识别。首先采用基于Gabor滤波器和Gaussian滤波器相结合的检测算法从视频中提取时空兴趣点,然后抽取兴趣点的静态特征、运动特征和时空特征,并分别对运动进行表征,最后利用基于时空码本的动作分类器对动作进行分类识别。在Weizmann和KTH两个行为数据集进行了测试,实验结果表明：基于时空特征的运动表示能够更好地适应摄像头移动、光照变化以及施动者的穿着和动作差异等环境因素的影响,取得更好的识别效果。     

基于视频深度学习的时空双流人物动作识别模型

深度学习在人物动作识别方面已取得较好的成效,但当前仍然需要充分利用视频中人物的外形信息和运动信息。为利用视频中的空间信息和时间信息来识别人物行为动作,提出一种时空双流视频人物动作识别模型。该模型首先利用两个卷积神经网络分别抽取视频动作片段空间和时间特征,接着融合这两个卷积神经网络并提取中层时空特征,最后将提取的中层特征输入到3D卷积神经网络来完成视频中人物动作的识别。在数据集UCF101和HMDB51上,进行视频人物动作识别实验。实验结果表明,所提出的基于时空双流的3D卷积神经网络模型能够有效地识别视频人物动作。     

时序增强的视频动作识别方法

针对视频动作识别中的时空建模问题,在深度学习框架下提出基于融合时空特征的时序增强动作识别方法.首先对输入视频应用稀疏时序采样策略,适应视频时长变化,降低视频级别时序建模成本.在识别阶段计算相邻特征图间的时序差异,以差异计算结果增强特征级别的运动信息.最后,利用残差结构与时序增强结构的组合方式提升网络整体时空建模能力.实验表明,文中算法在UCF101、HMDB51数据集上取得较高准确率,并在实际工业操作动作识别场景下,以较小的网络规模达到较优的识别效果.     

基于尺度自适应局部时空特征的足球比赛视频中的多运动员行为表示

为提高足球比赛视频中的多运动员行为识别的准确率,提出一种基于尺度自适应局部时空特征的足球比赛视频中的多运动员行为表示方法,利用时空兴趣点来表示足球比赛视频中的多运动员行为。首先将足球比赛视频序列中的多运动员行为看作是三维空间中的时空兴趣点的集合,然后采用直方图量化技术将时空兴趣点集合量化为维数固定的直方图（即时空单词）,最后采用K-means聚类算法生成时空码本。在聚类生成码本之前,对每个时空兴趣点都进行了归一化,以保证其缩放和平移不变性。实验结果表明,该方法能够大大减少足球比赛视频中的多运动员行为识别算法的计算量,显著提高识别的准确率。     

Action recognition by learning temporal slowness invariant features

Deep learning approaches emphasized on learning spatio-temporal features for action recognition. Different to previous works, we separate the spatio-temporal feature learning unity into the spatial feature learning and the spatial/temporal feature pooling procedures. Using the temporal slowness regularized independent subspace analysis network, we learn invariant spatial features from sampled video cubes. To be robust to the cluttered backgrounds, we incorporate the denoising criterion to our network. The local spatio-temporal features are obtained by pooling features from the spatial and the temporal aspects. The key points are that we learn spatial features from video cubes and pool features from spatial feature sequences. We evaluate the learned local spatio-temporal features on three benchmark action datasets. Extensive experiments demonstrate the effectiveness of the novel feature learning architecture.     

基于改进的深度卷积神经网络的人体动作识别方法

针对现有的动作识别算法的特征提取复杂、识别率低等问题,提出了基于批归一化变换(batch normalization)与GoogLeNet网络模型相结合的网络结构,将图像分类领域的批归一化思想应用到动作识别领域中进行训练算法改进,实现了对视频动作训练样本的网络输入进行微批量(mini-batch)归一化处理。该方法以RGB图像作为空间网络的输入,光流场作为时间网络输入,然后融合时空网络得到最终动作识别结果。在UCF101和HMDB51数据集上进行实验,分别取得了93.50%和68.32%的准确率。实验结果表明,改进的网络架构在视频人体动作识别问题上具有较高的识别准确率。     

(2+1)D多时空信息融合模型及在行为识别的应用

针对常规的卷积神经网络时空感受野尺度单一,难以提取视频中多变的时空信息的问题,利用（2+1）D模型将时间信息和空间信息在一定程度上解耦的特性,提出了（2+1）D多时空信息融合的卷积残差神经网络,并用于人体行为识别.该模型以3×3空间感受野为主,1×1空间感受野为辅,与3种不同时域感受野交叉组合构建了6种不同尺度的时空感受野.提出的多时空感受野融合模型能够同时获取不同尺度的时空信息,提取更丰富的人体行为特征,因此能够更有效识别不同时间周期、不同动作幅度的人体行为.另外提出了一种视频时序扩充方法,该方法能够同时在空间信息和时间序列扩充视频数据集,丰富训练样本.提出的方法在公共视频人体行为数据集UCF101和HMDB51上子视频的识别率超过或接近最新的视频行为识别方法.     

用于骨架行为识别的多维特征嵌合注意力机制

目的 在行为识别任务中,妥善利用时空建模与通道之间的相关性对于捕获丰富的动作信息至关重要。尽管图卷积网络在基于骨架信息的行为识别方面取得了稳步进展,但以往的注意力机制应用于图卷积网络时,其分类效果并未获得明显提升。基于兼顾时空交互与通道依赖关系的重要性,提出了多维特征嵌合注意力机制(multi-dimensional feature fusion attention mechanism,M2FA)。方法 不同于现今广泛应用的行为识别框架研究理念,如卷积块注意力模块(convolutional block attention module,CBAM)、双流自适应图卷积网络(two-stream adaptive graph convolutional network,2s-AGCN)等,M2FA通过嵌入在注意力机制框架中的特征融合模块显式地获取综合依赖信息。对于给定的特征图,M2FA沿着空间、时间和通道维度使用全局平均池化操作推断相应维度的特征描述符。特征图使用多维特征描述符的融合结果进行过滤学习以达到细化自适应特征的目的,并通过压缩全局动态信息的全局特征分支与仅使用逐点卷积层的局部特征分支相互嵌合获取多尺度动态信息。结果 实验在骨架行为识别数据集NTU-RGBD和Kinetics-Skeleton中进行,分析了M2FA与其基线方法2s-AGCN及最新提出的图卷积模型之间的识别准确率对比结果。在Kinetics-Skeleton验证集中,相比于基线方法2s-AGCN,M2FA分类准确率提高了1.8%;在NTU-RGBD的两个不同基准分支中,M2FA的分类准确率比基线方法2s-AGCN分别提高了1.6%和1.0%。同时,消融实验验证了多维特征嵌合机制的有效性。实验结果表明,提出的M2FA改善了图卷积骨架行为识别方法的分类效果。结论 通过与基线方法2s-AGCN及目前主流图卷积模型比较,多维特征嵌合注意力机制获得了最高的识别精度,可以集成至基于骨架信息的体系结构中进行端到端的训练,使分类结果更加准确。     

基于时空关联度的视频车祸识别算法

为了从一大段车祸视频中快速获取车祸发生时的那一小段视频,从中准确识别车祸视频中车辆碰撞的局部信息,从而尽快通知相关部门采取营救措施,避免二次事故的发生,实现快速辅助救援,提出了时空结合的车祸识别方法。通过对车祸视频进行局部特征点的提取和描述,以及对车祸视频进行局部特征时空角点检测,从而实现对车祸视频进行时空特征的提取。经特征提取后可以获得车祸视频的3DSIFT、STIP和3DHOG三种特征,对这三种特征分别进行不同组合的串联融合,融合后可以得到5种串联结果,对其串联特征进行K-MEANS聚类和KNN识别,从而实现车祸视频的识别。实验结果表明,利用时空关联的算法进行识别,能够准确找到车祸发生的片段,可以提高车祸识别精度。对比传统算法所得到的33%识别率,基于时空关联度的视频车祸识别算法可提高车祸识别精度至62%。     

基于特征传播的时域分割网络行为识别

为了高效、准确地获得视频中的行为类别和运动信息,减少计算的复杂度,文中提出一种融合特征传播和时域分割网络的视频行为识别算法.首先将视频分为3个小片段,分别从相应片段中提取关键帧,从而实现对长时间视频的建模;然后设计一个包含特征传播表观信息流和FlowNet运动信息流的改进时域分割网络(P-TSN),分别以RGB关键帧、RGB非关键帧、光流图为输入提取视频的表观信息流和运动信息流;最后将改进时域分割网络的BN-Inception描述子进行平均加权融合后送入Softmax层进行行为识别.在UCF101和HMDB51这2个数据集上分别取得了94.6%和69.4%的识别准确率,表明该算法能够有效地获得视频中空域表观信息和时域运动信息,提高了视频行为识别的准确率.     

Exploring probabilistic localized video representation for human action recognition

In recent years, the bag-of-words (BoW) video representations have achieved promising results in human action recognition in videos. By vector quantizing local spatial temporal (ST) features, the BoW video representation brings in simplicity and efficiency, but limitations too. First, the discretization of feature space in BoW inevitably results in ambiguity and information loss in video representation. Second, there exists no universal codebook for BoW representation. The codebook needs to be re-built when video corpus is changed. To tackle these issues, this paper explores a localized, continuous and probabilistic video representation. Specifically, the proposed representation encodes the visual and motion information of an ensemble of local ST features of a video into a distribution estimated by a generative probabilistic model. Furthermore, the probabilistic video representation naturally gives rise to an information-theoretic distance metric of videos. This makes the representation readily applicable to most discriminative classifiers, such as the nearest neighbor schemes and the kernel based classifiers. Experiments on two datasets, KTH and UCF sports, show that the proposed approach could deliver promising results.     

Video retrieval of near-duplicates using κ-nearest neighbor retrieval of spatio-temporal descriptors

This paper describes a novel methodology for implementing video search functions such as retrieval of near-duplicate videos and recognition of actions in surveillance video. Videos are divided into half-second clips whose stacked frames produce 3D space-time volumes of pixels. Pixel regions with consistent color and motion properties are extracted from these 3D volumes by a threshold-free hierarchical space-time segmentation technique. Each region is then described by a high-dimensional point whose components represent the position, orientation and, when possible, color of the region. In the indexing phase for a video database, these points are assigned labels that specify their video clip of origin. All the labeled points for all the clips are stored into a single binary tree for efficient -nearest neighbor retrieval. The retrieval phase uses video segments as queries. Half-second clips of these queries are again segmented by space-time segmentation to produce sets of points, and for each point the labels of its nearest neighbors are retrieved. The labels that receive the largest numbers of votes correspond to the database clips that are the most similar to the query video segment. We illustrate this approach for video indexing and retrieval and for action recognition. First, we describe retrieval experiments for dynamic logos, and for video queries that differ from the indexed broadcasts by the addition of large overlays. Then we describe experiments in which office actions (such as pulling and closing drawers, taking and storing items, picking up and putting down a phone) are recognized. Color information is ignored to insure independence of action recognition to people's appearance. One of the distinct advantages of using this approach for action recognition is that there is no need for detection or recognition of body parts.     

一种分步的融合时空信息的背景建模

自然场景中的光照突变和树枝、水面等不规则运动是背景建模的主要困难. 针对该问题,提出一种分步的融合时域信息和空域信息的背景建模方法. 在时域,采用具有光照不变性的颜色空间表征时域信息,并提出对噪声和光照突变具有较好适应性的码字聚类准则和自适应背景更新策略,构造了对噪声和光照突变具有较好适应性的时域信息背景模型. 在空域,通过采样将测试序列图像分成两幅子图,而后利用时域模型检测其中一幅子图,并将检测结果作为另一幅子图的先验信息,同时采用马尔科夫随机场（Markov random field,MRF）对其加以约束,最终检测其状态. 在多个测试视频序列上的实验结果表明,本文背景模型对于自然场景中的光照突变和不规则运动具有较好的适应性.     

人体动作的超兴趣点特征表述及识别

提出一种基于超兴趣点的动作特征描述方法,用于人体动作的识别。兴趣点特征描述了人体动作时变化显著的局部点信息,但其最大的缺陷在于离散的兴趣点间缺乏时间和空间上的结构关联。提出根据兴趣点间的时空距离,使用广度优先搜索邻居算法,将时空距离相近的兴趣点聚合成超兴趣点,该结构作为一个整体,反映人肢体在一定时空范围内的动作变化特征。与现有的基于局部兴趣点的动作识别算法相比,本文算法增加了兴趣点间的整体时空结构关系,提高了特征的区分度。实验采用两层分类方法对超兴趣点特征分类,实验结果表明该算法具有较好的识别率。     

T-STAM:基于双流时空注意力机制的端到端的动作识别模型

针对双流法进行视频动作识别时忽略特征通道间的相互联系、特征存在大量冗余的时空信息等问题,提出一种基于双流时空注意力机制的端到端的动作识别模型T-STAM,实现了对视频关键时空信息的充分利用。首先,将通道注意力机制引入到双流基础网络中,通过对特征通道间的依赖关系进行建模来校准通道信息,提高特征的表达能力。其次,提出一种基于CNN的时间注意力模型,使用较少的参数学习每帧的注意力得分,重点关注运动幅度明显的帧。同时提出一种多空间注意力模型,从不同角度计算每帧中各个位置的注意力得分,提取多个运动显著区域,并且对时空特征进行融合进一步增强视频的特征表示。最后,将融合后的特征输入到分类网络,按不同权重融合两流输出得到动作识别结果。在数据集HMDB51和UCF101上的实验结果表明T-STAM能有效地识别视频中的动作。     

基于动作图的视角无关动作识别

针对视角无关的动作识别,提出加权字典向量描述方法和动作图识别模型.将视频中的局部兴趣点特征和全局形状描述有机结合,形成加权字典向量的描述方法,该方法既具有兴趣点抗噪声强的优点,又可克服兴趣点无法识别静态动作的缺点.根据运动捕获、点云等三维运动数据构建能量曲线,提取关键姿势,生成基本运动单元,并通过自连接、向前连接和向后连接3种连接方式构成有向图,称为本质图.本质图向各个方向投影,根据节点近邻规则建立的有向图称为动作图.通过Na?ve Bayes训练动作图模型,采用Viterbi算法计算视频与动作图的匹配度,根据最大匹配度标定视频序列.动作图具有多角度投影和投影平滑过渡等特点,因此可识别任意角度、任意运动方向的视频序列.实验结果表明,该算法具有较好的识别效果,可识别单目视频、多目视频和多动作视频.     

基于时空灰度序特征的视频片段定位算法

基于灰度序特征的视频片段定位算法是解决视频片段定位问题的典型算法.这类算法存在的不足是：特征的唯一性表示能力不够,使得在召回率较高的情况下,定位检索的精度下降得较快;二次多项式级的时间复杂度使得响应时间过长,并对查询视频长度敏感.针对上述两个问题,提出了一种基于时空灰度序特征的视频片段定位算法,其关键步骤包括：(1) 在精确定位之前,通过引入线性时间复杂度的基于时空二值模式直方图特征(spatio-temporal binary pattern histogram,简称STBPH)的实时过滤算法以及基于二值时间灰度序特征(binarytemporal ordinal measure,简称BTOM)的快速过滤算法,大幅度减少精确定位阶段需要进行比较的候选视频片段个数;(2) 在精确定位阶段,通过引入唯一性表示能力更好且保持了较好鲁棒性的时空统一灰度序特征(jointspatio-temporal ordinal measure,简称JSTOM)进行序列匹配,显著提高了定位检索的精度.实验结果表明,该算法能够快速、准确地进行视频片段定位,大幅降低了对查询视频长度的敏感度.     

Spatio-Temporal Scale Selection in Video Data

This work presents a theory and methodology for simultaneous detection of local spatial and temporal scales in video data. The underlying idea is that if we process video data by spatio-temporal receptive fields at multiple spatial and temporal scales, we would like to generate hypotheses about the spatial extent and the temporal duration of the underlying spatio-temporal image structures that gave rise to the feature responses. For two types of spatio-temporal scale-space representations, (i) a non-causal Gaussian spatio-temporal scale space for offline analysis of pre-recorded video sequences and (ii) a time-causal and time-recursive spatio-temporal scale space for online analysis of real-time video streams, we express sufficient conditions for spatio-temporal feature detectors in terms of spatio-temporal receptive fields to deliver scale-covariant and scale-invariant feature responses. We present an in-depth theoretical analysis of the scale selection properties of eight types of spatio-temporal interest point detectors in terms of either: (i)–(ii) the spatial Laplacian applied to the first- and second-order temporal derivatives, (iii)–(iv) the determinant of the spatial Hessian applied to the first- and second-order temporal derivatives, (v) the determinant of the spatio-temporal Hessian matrix, (vi) the spatio-temporal Laplacian and (vii)–(viii) the first- and second-order temporal derivatives of the determinant of the spatial Hessian matrix. It is shown that seven of these spatio-temporal feature detectors allow for provable scale covariance and scale invariance. Then, we describe a time-causal and time-recursive algorithm for detecting sparse spatio-temporal interest points from video streams and show that it leads to intuitively reasonable results. An experimental quantification of the accuracy of the spatio-temporal scale estimates and the amount of temporal delay obtained from these spatio-temporal interest point detectors is given, showing that: (i) the spatial and temporal scale selection properties predicted by the continuous theory are well preserved in the discrete implementation and (ii) the spatial Laplacian or the determinant of the spatial Hessian applied to the first- and second-order temporal derivatives leads to much shorter temporal delays in a time-causal implementation compared to the determinant of the spatio-temporal Hessian or the first- and second-order temporal derivatives of the determinant of the spatial Hessian matrix.     

面向海量交通数据的HBase时空索引

针对HBase无法直接建立时空索引所带来的交通数据查询性能问题,基于HBase行键设计了面向海量交通数据的HBase时空索引。首先利用Geohash降维方法将二维空间位置数据转化为一维编码,再与时间维度进行组合;然后根据组合顺序的不同,提出了四种结构模型,分别讨论了模型的具体构成以及交通数据查询中的适应面;最后提出了相应的时空索引管理算法及基于Hbase时空索引的交通数据查询方法。通过实验验证了提出的HBase时空索引结构能有效提升海量交通数据的区域查询性能,并比较了四种时空索引结构在不同数据规模、不同查询半径以及不同时间范围的查询性能,量化验证了不同索引结构在交通数据查询中的适应场景。