基于ConvGRU和注意力特征融合的人体动作识别

在动作识别任务中,如何充分学习和利用视频的空间特征和时序特征的相关性,对最终识别结果尤为重要。针对传统动作识别方法忽略时空特征相关性及细小特征,导致识别精度下降的问题,本文提出了一种基于卷积门控循环单元(convolutional GRU, ConvGRU)和注意力特征融合(attentional feature fusion,AFF) 的人体动作识别方法。首先,使用Xception网络获取视频帧的空间特征提取网络,并引入时空激励(spatial-temporal excitation,STE) 模块和通道激励(channel excitation,CE) 模块,获取空间特征的同时加强时序动作的建模能力。此外,将传统的长短时记忆网络(long short term memory, LSTM)网络替换为ConvGRU网络,在提取时序特征的同时,利用卷积进一步挖掘视频帧的空间特征。最后,对输出分类器进行改进,引入基于改进的多尺度通道注意力的特征融合(MCAM-AFF)模块,加强对细小特征的识别能力,提升模型的准确率。实验结果表明:在UCF101数据集和HMDB51数据集上分别达到了95.66%和69.82%的识别准确率。该算法获取了更加完整的时空特征,与当前主流模型相比更具优越性。     

基于特征融合的人体行为识别

为克服单个行为表达方法有效性上的不足,提出了一种基于多特征融合和支持向量机(SVM)的人体行为识别(HAR)方法。首先,利用背景差分提取运动显著区域;然后提取运动显著区域的剪影直方图和光流直方图,并采取一定的融合策略,构建融合特征结合SVM识别人体行为。实验以广泛使用的公开数据集Weizmann为研究对象,正确识别率达到99.8%以上。结果表明,提出的特征融合及识别方法能有效地对人体行为进行识别;而且,由于规避了比较耗时的序列匹配操作,减少了计算量。     

基于关键点梯度特征描述的人体行为识别算法

提出一种新的人体行为识别方案并进行了算法实现。通过对视频序列在空间上高斯滤波,在时间轴向上Gabor滤波,提取出视频序列的关键点,对每个关键点邻域20×20的区域使用梯度位置朝向直方图进行描述,描述的序列可以表征视频序列的特征。与其他人体行为识别算法比较,不需要标记特定的特征区域和比较耗时的聚类算法,构建单个支持向量分类器即可达到好的识别率,算法简单有效。     

基于特征融合卷积神经网络的FMCW雷达人体动作识别

针对微多普勒特征识别人体动作的局限性,基于调频连续波( Frequency Modulated Continuous Wave,FMCW)雷达采用深度学习方法对人体动作识别,提出了一种特征融合卷积神经网络结构.利用FMCW雷达采样的人体动作回波数据分别构建出时间-距离特征和微多普勒特征图,将这两种特征图作为输入数据分别...     

Incomplete fingerprint recognition based on feature fusion and pattern entropy

Considering the inherent characteristics of incomplete fingerprint： local feature loss and global information distortion, the recognition progress has been mainly restricted by two critical problems： how to precisely extract informative features and still with compact representation of the incomplete fingerprint; and how to effectively measure the similarity between fingerprint images. In this paper, to handle the first problem, both the minutiae and orientation field feature are extracted and then fused to get a more comprehensive feature with scale and rotation invariability. Dealing with the second one, the pattern entropy is introduced to robustly measure the similarity of two incomplete fingerprints. Extensive experiments have been conducted on both those popular fingerprint databases and our extended databases containing more incomplete fingerprints. Meanwhile, thorough performance comparisons have been made with existing approaches. Experimental results show that our approach has more efficient ability especially in incomplete fingerprint recognition, and also performs well in both accuracy and efficiency.     

基于关键姿势的动作识别

人体动作识别是计算机视频和图像方面的一个热点问题,为了解决识别率不高、识别速度不快、不能实时识别,以及不同的人摆出相同动作时出现的识别误差,提出了一种能有效解决该问题的方法,该方法分析计算匹配视频帧序列,然后分类匹配后的视频帧,达到识别的目的。     

复杂环境下基于多特征决策融合的眼睛状态识别

针对常用图像特征容易受到复杂光照、头部运动等因素的影响导致眼状态识别算法的准确率降低的问题,本文在对多种红外条件下眼睛图像特征进行分析研究的基础上,选择具有旋转不变性和尺度不变性但对光照敏感的伪Zernike矩特征、简单并有效但对轮廓提取有较高要求的复杂度特征和对光照不敏感但容易受到头部运动影响的HOG特征作为眼状态识别的特征,提出了一种基于多特征决策融合的眼状态识别算法。首先建立上述3种特征相应的支持向量机(SVM)分类器,然后利用自动权值学习算法得到3个特征分类器的决策权重,最后综合利用不同特征的性能特点对3个分类器的识别结果进行决策融合从而得到最终识别结果,提高了眼状态识别算法的鲁棒性。实验结果表明,本文算法能够较好克服光照和头部运动对眼睛状态识别的影响,识别准确率达到91.9%。     

Multiple feature fusion for unimodal emotion recognition

A new semi-serial fusion method of multiple feature based on learning using privileged information(LUPI) model was put forward.The exploitation of LUPI paradigm permits the improvement of the learning accuracy and its stability,by additional information and computations using optimization methods.The execution time is also reduced,by sparsity and dimension of testing feature.The essence of improvements obtained using multiple features types for the emotion recognition(speech expression recognition),is particularly applicable when there is only one modality but still need to improve the recognition.The results show that the LUPI in unimodal case is effective when the size of the feature is considerable.In comparison to other methods using one type of features or combining them in a concatenated way,this new method outperforms others in recognition accuracy,execution reduction,and stability.     

基于LSTM网络和特征融合的通信干扰识别

针对现有通信干扰信号识别方法识别效果不佳的问题,提出了一种基于长短时记忆网络(Long Short-Term Memory,LSTM)和特征融合的通信干扰识别方法.该方法利用LSTM网络提取干扰信号的特征,通过LSTM强大的序列特征提取能力提升干扰信号特征提取的性能;通过提取信号的时域和频域特征后进行特征融合,使用全连...     

基于Kinect和金字塔特征的行为识别算法

提出了一种基于Kinect和金字塔特征的行为识别算法。在算法中,Kinect不仅能够获得RGB信息,还能获得与RGB信息对应的深度信息;而金字塔特征不仅描述了人体行为的全局形状和局部细节信息,而且还描述了人体行为的空间信息。通过不同核函数的支持向量机(SVM)分类器在具有挑战性的DHA数据集的试验结果表明,金字塔特征在RGB和深度图上都能获得令人满意的性能,且当深度特征和RGB特征融合时,其性能获得了进一步的提高,识别率达到96.2%,远高于一些具有代表性的行为描述子。     

Human action recognition based on convolutional neural network and spatial pyramid representation

Detecting and recognizing human action in natural scenarios, such as indoor and outdoor, is a significant technique in computer vision and intelligent systems, which is widely applied in video surveillance, pedestrian tracking and human-computer interaction. Conventional approaches have been proposed based on various features and achieved impressive performance. However, these methods failed to cope with partial occlusion and changes of posture. In order to address these limitations, we propose a novel human action recognition method. More specifically, in order to capture image spatial composition, we leverage a three-level spatial pyramid feature extraction scheme, where each pyramid is encoded by local features. Thereafter, regions generated by a proposal algorithm are fed into a dual-aggregation net for deep representation extraction. Afterwards, both local features and deep features are fused to describe each image. To describe human action category, we design a metric CXQDA based on Cosine measure and Cross-view Quadratic Discriminant Analysis (XQDA) to calculate the similarity among different action categories. Experimental results demonstrate that our proposed method can effectively cope with object scale variations, partial occlusion and achieve competitive performance.     

Deeply fusing multi-model quality-aware features for sophisticated human activity understanding

Detecting and understanding human action under sophisticated lighting condition and backgrounds, also known as human action recognition in real-world context, is an indispensable component in modern intelligent systems and has becoming a hot research topic currently. Nowadays, human action recognition is still a tough challenge due to intra-class and inter-class, environment and temporal-level differences of the same action. Algorithms based on the single visual channel cannot achieve satisfactory performance. Thus, in this paper, we propose a novel action recognition framework towards sophisticated activity understanding, focusing on intelligently combining multimodel quality-related action features. Specifically, we first design a multi-channel feature fusion (MCFF) algorithm to capture visual appearance, motion and acoustic patterns from each video frame, where image-level labels are characterized by choosing high quality multimodel features. Subsequently, we design an adaptive key frame selection algorithm that can be applied to characterize human action from human action video stream. Thereafter, we engineer a multimodel feature based on an auxiliary human action retrieval system to achieve sophisticated activity understanding. Extensive experimental evaluations have demonstrated that the effectiveness and robustness of our proposed method.     

基于多时空特征的人体动作识别算法

针对基于单时空特征的人体动作识算法的不足,提 出了一种基于多时空特征的人体动 作识别算法。通过在KTH与YouTube action公共动作数据集上的实验表明,本文提出的多时空特征的动作识别算法在较小码书的 情况下,具有 较好的区分性、鲁棒性以及实时性,且比一些且具有代表性的算法性能更好。     

基于稠密轨迹特征的红外人体行为识别

提出了一种使用基于稠密轨迹(DT)融合特征的红外人体行为识别(HAR)方法。主要流程如下:1)通过稠密采样获得输入行为视频的DT;2)计算DT的方向梯度直方图(HOG)、光流直方图(HOF)和运动边界描述子(MBH)3个描述子;3)基于DT的HOG、HOF和MBH,并采取词袋库模型和一定的融合策略,构建融合特征;4)以第3步所构建的融合特征为k近邻分类器(k-NN)的输入,完成人体HAR。实验以IADB红外行为库为研究对象,正确识别率达到96.7%。结果表明,提出的特征融合及识别方法能有效地对红外人体行为进行识别。     

多分类器融合的指纹全局特征协同识别

指纹识别是生物特征识别中的热点,指纹全局特征识别具有明显的优势,但是单一分类器一般不能取得满意的识别效果。本文采用贝叶斯理论分析了常见的积、和、中值以及投票多分类器融合方法,并根据实际的选举情形,对投票法进行了2种改进。然后对3种指纹全局特征协同识别分类器:灰度值、主分量以及方向场分类器进行决策层融合,并采用了一种崭新而高效的协同模式识别方法。对FVC2002指纹库的实验表明:该方法具有较好的分类性能,预处理与特征提取简单、计算复杂度低、识别速度快、对污损指纹具有可靠的识别率、鲁棒性强,而且应用于身份认证中也取得了较好的认证效果。     

基于区域矩的舰船图像目标特征融合方法研究

为了增强特征对图像目标的描述能力,一类重要的方法是将不同类型的特征融合。面向舰船图像目标识别,研究了基于区域矩的特征融合方法。区域矩不是计算原始图像像素的矩,而是计算图像特征的矩,具体是首先设计合适的目标特征,然后利用中心矩、Hu氏不变矩和径向矩实现特征融合,并基于KNN分类实现目标识别。仿真实验表明,与区域协方差相比,区域矩可以更好地融合舰船可见光图像或红外图像的目标特征,而且没有随机初始化问题,所以实用性和有效性更高。     

基于多特征融合的弱小运动目标识别

利用单个特征识别强噪声中的弱小运动目标,常因所提取的目标特征与噪声特征易混淆而导致高的虚警率.提出一种新的基于多特征融合的弱小运动目标识别方法.分析了弱小运动目标的连续相关性、面积及质心位置偏移这三个特征的可靠性及提取方法,对获取的特征值进行归一化后采用多特征融合的方法构造更具有鲁棒性的联合特征,确定了以具有最大多特征融合值为真实目标的决策方法.通过与采用单一特征的目标识别方法进行比较,证明了提出的多特征融合方法能更准确地识别弱小运动目标.     

Cattle behavior recognition based on feature fusion under a dual attention mechanism

In recent years, artificial intelligence has been widely used in such fields as agricultural informatization, precision agriculture and precision animal husbandry. Due to limited research on deep learning in real-time agricultural and pastoral situations, deep learning and computer vision have become very important topics in the agricultural field. Recent studies have shown that the fusion of features under different attention mechanisms will help advance the utilization of such features, and will thus influence the accuracy and generalization ability of the models used. In this paper, we propose a lightweight network structure based on feature fusion under a dual attention mechanism with the same activation and joint loss functions. More specifically, we propose an innovative method to improve the network structure of two different attention mechanisms, and achieve feature fusion by combining the two. At the same time, we keep the activation functions consistent with those of the original network structure, and we develop a joint loss function to expand the use of various features. We also take the novel approach of applying the trajectory behavior analysis method to walking and standing. Experiments using both a publicly available data set and a data set obtained from a farm show that our algorithm achieves state-of-the-art performance in terms of accuracy and generalization ability, as compared to other methods.     

基于总错误率和特征关联的自适应融合多模态生物特征识别

针对单生物特征识别准确率和鲁棒性差的问题, 提出了一种基于总错误率(TER)和特征关联自适应融合多模态生物特 征识别方法。首先将TER作为判别特征引入到多模态识别,以代替传统的匹配分 数;其次在不确定度量理论的基 础上,考虑人脸特征和语音特征之间的时空关联性,提出了一种基于特征关联的多特征 自适应融合策略,利用特征关联 系数自适应调节不同识别特征对识别结果的贡献。仿真实验表明,与几种代表性的融合算法 相比,本文所 提出的融合模式可以有效提高多生物特征识别系统的准确性和鲁棒性。     

基于HALCON与SURF的多特征融合条烟识别系统

烟草物流中心工作量大,枯燥单一,导致分拣过程中经常出现多拿、少拿以及错拿等错误分拣现象,这极大的影响了分拣的效率,甚至导致一些不必要的损失。针对此现象,设计了一种基于HALCON与SURF的多特征融合条烟识别系统。不但提出了新的条烟图像特征描述方法,同时针对分拣生产线的特点,对识别策略进行了相应的改进设计。实验表明,此方案具有较好的应用前景。