Research on Facial Fatigue Detection of Drivers with Multi-feature Fusion

In order to solve the shortcomings of current fatigue detection methods such as low accuracy or poor real-time performance,a fatigue detection method based on multi-feature fusion is proposed.Firstly,the HOG face detection algorithm and KCF target tracking algorithm are integrated and deformable convolutional neural network is introduced to identify the state of extracted eyes and mouth,fast track the detected faces and extract continuous and stable target faces for more efficient extraction.Then the head pose algorithm is introduced to detect the driver’s head in real time and obtain the driver’s head state information.Finally,a multi-feature fusion fatigue detection method is proposed based on the state of the eyes,mouth and head.According to the experimental results,the proposed method can detect the driver’s fatigue state in real time with high accuracy and good robustness compared with the current fatigue detection algorithms.

SA-Model:Multi-Feature Fusion Poetic Sentiment Analysis Based on a Hybrid Word Vector Model

Sentiment analysis in Chinese classical poetry has become a prominent topic in historical and cultural tracing,ancient literature research,etc.However,the existing research on sentiment analysis is relatively small.It does not effectively solve the problems such as the weak feature extraction ability of poetry text,which leads to the low performance of the model on sentiment analysis for Chinese classical poetry.In this research,we offer the SA-Model,a poetic sentiment analysis model.SA-Model firstly extracts text vector information and fuses it through Bidirectional encoder representation from transformers-Whole word masking-extension(BERT-wwmext)and Enhanced representation through knowledge integration(ERNIE)to enrich text vector information;Secondly,it incorporates numerous encoders to remove text features at multiple levels,thereby increasing text feature information,improving text semantics accuracy,and enhancing the model’s learning and generalization capabilities;finally,multi-feature fusion poetry sentiment analysis model is constructed.The feasibility and accuracy of the model are validated through the ancient poetry sentiment corpus.Compared with other baseline models,the experimental findings indicate that SA-Model may increase the accuracy of text semantics and hence improve the capability of poetry sentiment analysis.

Multi-Feature Fusion Book Recommendation Model Based on Deep Neural Network

The traditional recommendation algorithm represented by the collaborative filtering algorithm is the most classical and widely recommended algorithm in the practical industry.Most book recommendation systems also use this algorithm.However,the traditional recommendation algorithm represented by the collaborative filtering algorithm cannot deal with the data sparsity well.This algorithm only uses the shallow feature design of the interaction between readers and books,so it fails to achieve the high-level abstract learning of the relevant attribute features of readers and books,leading to a decline in recommendation performance.Given the above problems,this study uses deep learning technology to model readers’book borrowing probability.It builds a recommendation system model through themulti-layer neural network and inputs the features extracted from readers and books into the network,and then profoundly integrates the features of readers and books through the multi-layer neural network.The hidden deep interaction between readers and books is explored accordingly.Thus,the quality of book recommendation performance will be significantly improved.In the experiment,the evaluation indexes ofHR@10,MRR,andNDCGof the deep neural network recommendation model constructed in this paper are higher than those of the traditional recommendation algorithm,which verifies the effectiveness of the model in the book recommendation.

The detection method of low-rate DoS attack based on multi-feature fusion

As a new type of Denial of Service(DoS)attacks,the Low-rate Denial of Service(LDoS)attacks make the traditional method of detecting Distributed Denial of Service Attack(DDoS)attacks useless due to the characteristics of a low average rate and concealment.With features extracted from the network traffic,a new detection approach based on multi-feature fusion is proposed to solve the problem in this paper.An attack feature set containing the Acknowledge character(ACK)sequence number,the packet size,and the queue length is used to classify normal and LDoS attack traffics.Each feature is digitalized and preprocessed to fit the input of the K-Nearest Neighbor(KNN)classifier separately,and to obtain the decision contour matrix.Then a posteriori probability in the matrix is fused,and the fusion decision index D is used as the basis of detecting the LDoS attacks.Experiments proved that the detection rate of the multi-feature fusion algorithm is higher than those of the single-based detection method and other algorithms.

Smoke root detection from video sequences based on multi-feature fusion

Smoke detection is the most commonly used method in early warning of fire and is widely used in forest detection.Most existing smoke detection methods contain empty spaces and obstacles which interfere with detection and extract false smoke roots.This study developed a new smoke roots search algorithm based on a multi-feature fusion dynamic extraction strategy.This determines smoke origin candidate points and region based on a multi-frame discrete confidence level.The results show that the new method provides a more complete smoke contour with no background interference,compared to the results using existing methods.Unlike video-based methods that rely on continuous frames,an adaptive threshold method was developed to build the judgment image set composed of non-consecutive frames.The smoke roots origin search algorithm increased the detection rate and significantly reduced false detection rate compared to existing methods.

Multi-Feature Fusion-Guided Multiscale Bidirectional Attention Networks for Logistics Pallet Segmentation

In the smart logistics industry,unmanned forklifts that intelligently identify logistics pallets can improve work efficiency in warehousing and transportation and are better than traditional manual forklifts driven by humans.Therefore,they play a critical role in smart warehousing,and semantics segmentation is an effective method to realize the intelligent identification of logistics pallets.However,most current recognition algorithms are ineffective due to the diverse types of pallets,their complex shapes,frequent blockades in production environments,and changing lighting conditions.This paper proposes a novel multi-feature fusion-guided multiscale bidirectional attention(MFMBA)neural network for logistics pallet segmentation.To better predict the foreground category(the pallet)and the background category(the cargo)of a pallet image,our approach extracts three types of features(grayscale,texture,and Hue,Saturation,Value features)and fuses them.The multiscale architecture deals with the problem that the size and shape of the pallet may appear different in the image in the actual,complex environment,which usually makes feature extraction difficult.Our study proposes a multiscale architecture that can extract additional semantic features.Also,since a traditional attention mechanism only assigns attention rights from a single direction,we designed a bidirectional attention mechanism that assigns cross-attention weights to each feature from two directions,horizontally and vertically,significantly improving segmentation.Finally,comparative experimental results show that the precision of the proposed algorithm is 0.53%–8.77%better than that of other methods we compared.

Hierarchical particle filter tracking algorithm based on multi-feature fusion

A hierarchical particle filter（HPF） framework based on multi-feature fusion is proposed.The proposed HPF effectively uses different feature information to avoid the tracking failure based on the single feature in a complicated environment.In this approach,the Harris algorithm is introduced to detect the corner points of the object,and the corner matching algorithm based on singular value decomposition is used to compute the firstorder weights and make particles centralize in the high likelihood area.Then the local binary pattern（LBP） operator is used to build the observation model of the target based on the color and texture features,by which the second-order weights of particles and the accurate location of the target can be obtained.Moreover,a backstepping controller is proposed to complete the whole tracking system.Simulations and experiments are carried out,and the results show that the HPF algorithm with the backstepping controller achieves stable and accurate tracking with good robustness in complex environments.

Medical image fusion based on pulse coupled neural networks and multi-feature fuzzy clustering

Medical image fusion plays an important role in clinical applications such as image-guided surgery, image-guided radiotherapy, noninvasive diagnosis, and treatment planning. In order to retain useful information and get more reliable results, a novel medical image fusion algorithm based on pulse coupled neural networks (PCNN) and multi-feature fuzzy clustering is proposed, which makes use of the multi-feature of image and combines the advantages of the local entropy and variance of local entropy based PCNN. The results of experiments indicate that the proposed image fusion method can better preserve the image details and robustness and significantly improve the image visual effect than the other fusion methods with less information distortion.

Multi-Feature Fusion Based Relative Pose Adaptive Estimation for On-Orbit Servicing of Non-Cooperative Spacecraft

On-orbit servicing, such as spacecraft maintenance, on-orbit assembly, refueling, and de-orbiting, can reduce the cost of space missions, improve the performance of spacecraft, and extend its life span. The relative state between the servicing and target spacecraft is vital for on-orbit servicing missions, especially the final approaching stage. The major challenge of this stage is that the observed features of the target are incomplete or are constantly changing due to the short distance and limited Field of View (FOV) of camera. Different from cooperative spacecraft, non-cooperative target does not have artificial feature markers. Therefore, contour features, including triangle supports of solar array, docking ring, and corner points of the spacecraft body, are used as the measuring features. To overcome the drawback of FOV limitation and imaging ambiguity of the camera, a "selfie stick" structure and a self-calibration strategy were implemented, ensuring that part of the contour features could be observed precisely when the two spacecraft approached each other. The observed features were constantly changing as the relative distance shortened. It was difficult to build a unified measurement model for different types of features, including points, line segments, and circle. Therefore, dual quaternion was implemented to model the relative dynamics and measuring features. With the consideration of state uncertainty of the target, a fuzzy adaptive strong tracking filter( FASTF) combining fuzzy logic adaptive controller (FLAC) with strong tracking filter(STF) was designed to robustly estimate the relative states between the servicing spacecraft and the target. Finally, the effectiveness of the strategy was verified by mathematical simulation. The achievement of this research provides a theoretical and technical foundation for future on-orbit servicing missions.

A Multi-feature Fusion Apple Classification Method Based on Image Processing and Improved SVM

In order to achieve accurate classification of apple, a multi-feature fusion classification method based on image processing and improved SVM was proposed in this paper. The method was mainly divided into four parts, including image preprocessing, background segmentation, feature extraction and multi-feature fusion classification with improved SVM. Firstly, the homomorphic filtering algorithm was used to improve the quality of apple images. Secondly, the images were converted to HLS space. The background was segmented by the QTSU algorithm. Morphological processing was employed to remove fruit stem and surface defect areas. And apple contours were extracted with the Canny algorithm. Then, apples’ size, shape, color, defect and texture features were extracted. Finally, the cross verification method was used to optimize the penalty factor in SVM. A multi-feature fusion classification model was established. And the weight of each index was calculated by Fisher. In this study, 146 apple samples were selected for training and 61 apple samples were selected for testing. The test results showed that the accuracy of the classification method proposed in this paper was 96.72%, which can provide a reference for apple automatic classification.

A classification method of building structures based on multi-feature fusion of UAV remote sensing images

In order to improve the accuracy of building structure identification using remote sensing images,a building structure classification method based on multi-feature fusion of UAV remote sensing image is proposed in this paper.Three identification approaches of remote sensing images are integrated in this method:object-oriented,texture feature,and digital elevation based on DSM and DEM.So RGB threshold classification method is used to classify the identification results.The accuracy of building structure classification based on each feature and the multi-feature fusion are compared and analyzed.The results show that the building structure classification method is feasible and can accurately identify the structures in large-area remote sensing images.

改进FCENet的自然场景文本检测算法

针对自然场景文本检测中由于背景复杂、尺度多变、形状弯曲等造成的检测难题,提出了一种改进FCENet(Fourier contour embedding network)的场景文本检测算法。该算法基于FCENet并引入了多尺度残差特征增强模块和多尺度注意力特征融合模块。多尺度残差特征增强模块作为骨干网络顶层的残差分支,增强了特征金字塔结构自上而下的高层语义信息流动,提高了文本像素分类能力,有效减少误检现象。多尺度注意力特征融合模块使不同语义和尺度的特征能够更好地融合,结合自底向上的特征融合网络,有效避免文本过度分割并提高了弯曲文本的检测能力。实验结果表明,该方法在弯曲文本数据集CTW1500和Total-Text上的综合指标F值分别达到了86.2%和86.5%,相比原算法FCENet分别提升了1.1和0.7个百分点。

从“各行其是”到“同舟共济”:乡村情感治理的生成逻辑

随着社会转型升级,乡村由“熟人社会”过渡为“半熟人社会”,村民异质性凸显,旧有情感规则生存的社会基础发生巨变,原有社会规范式微。在理性选择下,集体行动的生长空间被严重压缩。尤其面对转型中的风险,显著地表现为村民行为原子化、响应服务被动化与资源储备薄弱化,消解了乡村治理的效力。如何推动村民从“各行其是”转向“同舟共济”,打造面向现代化的乡村治理共同体,是实现基层治理能力现代化目标面临的重要课题。通过对M村的考察,构建情感治理的分析框架,分析情感治理如何作用于基层治理体系。研究发现,“情感嵌入—情感融合—情感输出”能有效重塑个体情感,改善邻里关系;在情感共鸣中催生群体认同,促成情感联盟,凝聚集体力量。在集体联盟的感召下,乡土情怀升华为家国情怀,打造乡村治理共同体。“个体情感—群体情感—家国情感”相辅相成、螺旋式地演进,推动基层治理有效转型。

融合项目特征级信息的稀疏兴趣网络序列推荐

在以往提取多兴趣嵌入的序列推荐模型中仅能通过聚类的方法发现少量兴趣概念,忽视项目交互序列中特征级信息对最终推荐结果的影响。针对此问题,对传统的多兴趣序列推荐模型进行改进,提出一种融合项目特征级信息的稀疏兴趣网络序列推荐模型。实验结果表明,相比其它模型,该模型可以更好捕捉用户的多样化偏好并缓解冷启动问题。在给定数据集上,该模型比传统的序列推荐模型在命中率上平均提高了6.4%,归一化折损累计增益平均提高了8.7%。

位置标签增强的中文医学命名实体级联识别

针对一般领域的命名实体识别方法不能直接用于中文医学专业实体的识别,现有的相关研究只专注于英文文本和扁平结构的医学实体识别等问题,通过对专业领域实体识别方法的研究,结合中文医学实体的特点提出了一种面向中文医学实体的级联识别方法。将每个字符元素相对于实体的位置标签嵌入模型,并结合中文医学实体跨度内不同元素的重要程度进行实体的融合表示。通过序列标注方法检测字符的位置标签,利用字符的位置信息指导候选实体生成,并进行实体语义分类。模型在CMeEE和CCKS2018数据集以及中文糖尿病科研文献数据集上分别进行扁平实体、嵌套实体和不连续性长实体的识别实验。实验结果表明,该方法能够有效地识别中文医学文本中不同结构的实体。

多源知识图谱事件知识融合方法研究

以事件为中心的动态知识对事件预测等应用至关重要,但现有知识图谱主要关注以实体为中心的静态知识,难以满足需求。本文提出一种融合多源知识的高质量事件知识图谱构造方法,首先定义全局事件模式,利用标签类别从源知识图谱中提取事件知识并构造临时事件知识图,提出关系扩充规则对临时事件知识图进行扩充,改进实体对齐Attce模型,基于TransD模型对多个临时事件知识图进行联合嵌入学习,以提高实体对齐和冲突发现的效率;利用事件描述完整度计算源知识图谱可信度,发生冲突时作为判别标准进行处理。经过在真实数据集上的实验,验证了该方法的准确性和有效性。

骨架引导的多模态视频异常行为检测方法

视频异常行为检测是智能视频监控分析的一项重要且具有挑战性的任务,旨在自动发现异常事件。针对只采用单骨架模态导致部分相似运动模式的行为难以区分和缺乏时间全局信息的问题,提出骨架引导的多模态异常行为检测方法。为了充分利用RGB视频模态和骨架模态的优势进行相似行为下的异常行为检测,将从骨架模态中提取的动作行为特征作为引导,使用新的空间嵌入来加强RGB视频和骨架姿态之间的对应关系。同时使用时间自注意力提取相同节点的帧间关系,以捕获时间的全局信息,有效提取具有区分性的异常行为特征。在两个大型公开标准数据集上的实验结果表明所提方法能够有效加强骨架引导的多模态特征在空间和模态上的对应关系,并捕获时空图卷积缺乏的时间全局信息,使运动模式相似的异常行为实现更准确检测。

基于分层融合策略和上下文信息嵌入的多模态情绪识别

现有的多模态融合策略大多将不同模态特征进行简单拼接,忽略了针对单个模态固有特点的个性化融合需求。同时,在情绪识别阶段,独立地看待单个话语的情绪而不考虑其在前后话语语境下的情绪状态,可能导致情绪识别错误。为了解决上述问题,提出一种基于分层融合策略和上下文信息嵌入的多模态情绪识别方法,通过分层融合策略,采用层次递进的方式,依次融合不同的模态特征,以便减少单个模态的噪声干扰并解决不同模态间表达不一致的问题。该方法还充分利用融合后模态的上下文信息,综合考虑单个话语在上下文语境中的情绪表示,以便提升情绪识别的效果。在二分类情绪识别任务中,该方法的准确率比SOTA模型提升1.54%。在多分类情绪识别任务中,该方法的F1值比SOTA模型提升2.79%。

时频域多尺度交叉注意力融合的时间序列分类方法

针对时间序列子序列间的潜在信息交互不足导致分类准确率低的问题,提出时频域多尺度交叉注意力融合的时间序列分类方法TFFormer(Time-Frequency Transformer)。首先,将原始时间序列的时频域谱分别划分为等长子序列,经线性投影后加入位置信息解决时间序列的点值耦合问题;其次,通过改进的多头自注意力(IMHA)模块使模型关注更重要的序列特征,解决长时间序列的前后依赖问题;最后,构造多尺度时频域交叉注意力(CMA)模块增强时间序列在时域和频域之间的信息交互,使模型进一步挖掘序列的频域信息。实验结果表明,在Trace、StarLightCurves和UWaveGestureLibraryAll数据集上,相较于全卷积网络(FCN),所提方法的分类准确率分别提高了0.3、0.9和1.4个百分点,验证了通过增强时间序列时域和频域间的信息交互,可以提高模型收敛速度和分类精度。

面向多模态知识图谱的实体对齐方法研究

实体对齐是构建知识图谱的重要环节,也是该领域的一个研究热点.现有实体对齐工作在包含文本、图片的多模态知识图谱数据集DB15K-FB15K和YAGO15K-FB15K上做了大量研究,但是它们仅局限于文本和图片两种模态,且在多模态知识融合方面的性能并不显著.为弥补已有工作的不足,本文构建了一个包含文本、图片、视频的多模态知识图谱数据集Douban-Baidu,并提出了EA-MMKG模型来解决多模态知识图谱实体对齐问题.EA-MMKG包含两部分:多模态知识嵌入模块和多模态知识交互融合模块.具体来讲,多模态知识嵌入模块由关系三元组嵌入、图片嵌入、视频嵌入和属性三元组嵌入4个部分组成;多模态知识交互融合模块采用了基于注意力的融合机制来融合从文本、图片、视频3种模态中提取的特征信息,从而使得各模态之间的交互更加充分、融合效果更好,并最终提高多模态知识图谱实体对齐的性能.实验结果表明,EA-MMKG模型在Douban-Baidu数据集、DB15K-FB15K数据集和YAGO15K-FB15K数据集上的性能均优于现有的模型.