多形状局部区域神经网络结构的行人再识别

目的 目前,行人再识别领域将行人图像的全局和局部特征相结合的方法已经成为基本的解决方法。现有的基于局部特征的方法更多的是侧重于定位具有特定的语义区域,这样增加了学习难度,并且对于差异较大的图像场景不具有鲁棒性。为了解决上述问题,通过对网络结构进行改进提出一种多形状局部区域网络（MSPN）结构,它具有多分支并将横向和纵向条状的特征作为局部特征,能够端到端进行训练。方法 网络的多个分支设计可以同时获得多粒度和多形状的局部特征,其中一个分支表示全局特征的学习,两个分支表示横条状不同粒度的局部特征学习,最后一个分支表示竖条状局部特征学习。网络不再学习定位具有特定语义的区域,而是将图像提取的特征切分成横向和竖向的若干条作为局部特征。不同分支条的形状和数量不一致,最后获得不同粒度或不同形状的局部特征信息。因为切分方向的不同,多粒度多形状的局部特征缓解了行人在不同图像中无法对齐的问题。结果 在包括Market-1501、DukeMTMC-ReID和CUHK03在内的主流评估数据集上的综合实验表明,多形状局部区域神经网络和现有的主要方法相比具有更好的表现。其中在数据集Market-1501上达到84.57%的平均准确率（mAP）和94.51%的rank-1准确率。结论 多形状局部区域网络能够学习得到判别能力更强的深度学习模型,从而有效地提升行人再识别的准确率。     

A part-based attention network for person re-identification

Multimedia Tools and Applications - Person re-identification (re-id) is the task of recognizing images of the same pedestrian captured by different cameras with non-overlapping views. Person re-id...     

自适应感受野网络的行人重识别

行人重识别通常删除特征提取网络中的最后一个空间下采样操作,以增加最后输出特征图的分辨率,保留更多的细粒度特征.然而,这种操作会大幅减小神经网络的感受野,而更大的感受野可以为行人重识别提供更多的上下文信息.同时,在实际的视觉皮层中,相同区域的神经元的感受野是不同的,但当前行人重识别网络的设计大多忽视了这一点.为了解决上述问题,提出一种新颖的自适应感受野网络.网络的设计受启发于生物的视觉系统,通过在多分支网络上设置不同大小的感受野,结合注意力机制让网络自行选择合适的感受野特征,从而实现网络感受野的自适应,并且采用分组卷积使得自适应感受野模块更加轻量级.同时在各个分支利用空洞卷积增大感受野,补偿删除最后下采样操作所减少的网络感受野.在公开的大规模数据集上进行实验,实验结果表明,所提出的算法相比于基线方法有显著的提升,当使用ResNet-50作为特征提取网络时,在DukeMTMC-reID、Market-1501数据集上的Rank-1和mAP分别达到89.2%和76.0%、95.2%和87.2%.与现有方法相比,所提出算法在精度上有明显的提升.     

两路互补对称CNN结构的行人再识别

目的 行人再识别的任务是研究如何在海量监控数据中准确地识别出某个特定场合中曾经出现过的人,已成为公共安全领域中一项新的且具有挑战性的研究课题。其挑战在于,行人在图像中有较大的姿态、视角、光照等变化,这些复杂的变化会严重影响行人再识别性能。近年来,以卷积神经网络（CNN）为代表的深度学习方法在计算机视觉领域取得了巨大的成功,也带动了行人再识别领域的相关研究。CNN有效地克服了行人变化,取得较高的准确率。然而,由于行人再识别数据集中行人标注量小,利用现有的一路CNN模型,其训练过程并不够充分,影响了深度学习模型的鉴别能力。为了解决上述问题,通过对网络结构进行改进,提出一种两路互补对称的CNN结构用于行人再识别任务。方法 本文方法每次同时输入两路样本,其中每路样本之间具有互补特性,此时在有限的训练样本下,输入的组合会更加多样化,CNN模型的训练过程更加丰富。结果 对本文提出的方法在两个公开的大规模数据集（Market-1501和DukeMTMC-reID）上进行实验评估,相比于基线方法有稳定的提升,相比于现存的其他一些方法,其结果也有竞争力。在Market-1501数据集上,1选识别正确率和平均精度均值分别达到了73.25%和48.44%。在DukeMTMC-reID数据集上,1选识别正确率和平均精度均值分别达到了63.02%和41.15%。结论 本文提出的两路互补对称CNN结构的行人再识别方法,能够在现有的有限训练样本下,更加充分地训练CNN模型,学习得到鉴别能力更强的深度学习模型,从而有效地提升行人再识别的性能。     

A multi-branch attention and alignment network for person re-identification

Applied Intelligence - Person re-identification plays a critical role in video surveillance and has a variety of applications. However, the body misalignment caused by detectors or pose changes...     

基于视觉信息积累的行人重识别网络

在以往的行人重识别方法中,绝大部分的工作集中于图像注意力区域的学习,却忽视了非注意力区域对最终特征学习的影响,如果在关注图像注意力区域的同时加强非注意力区域的特征学习,可进一步丰富最终的行人特征,有利于行人身份信息的准确识别。基于此,提出了视觉信息积累网络(VIA Net),该网络整体采用两分支结构,一个分支倾向于学习图像的全局特征,另一个分支则拓展为多分支结构,通过结合注意力区域和非注意力区域的特征逐步加强局部特征的学习,实现视觉信息的积累,进一步丰富特征信息。实验结果表明,在Market-1501等行人重识别数据集上,所提出的VIA Net网络达到了较高的实验性能;同时,在In-Shop Clothes Retrieval数据集上的实验证明：该网络也适用于一般的图像检索任务,具有一定的通用性。     

基于增强特征融合网络的行人再识别

针对行人再识别问题,目前多数方法将行人的局部或全局特征分开考虑,从而忽略了行人整体之间的关系,即行人全局特征和局部特征之间的联系。本文提出一种增强特征融合网络(enhanced feature convergent network,EFCN)。在全局分支中,提出适用于获取全局特征的注意力网络作为嵌入特征,嵌入在基础网络模型中以提取行人的全局特征;在局部分支中,提出循环门单元变换网络(gated recurrent unit change network,GRU-CN)得到代表性的局部特征;再使用特征融合方法将全局特征和局部特征融合成最终的行人特征;最后借助损失函数训练网络。通过大量的对比实验表明,该算法网络模型在标准的Re-ID数据集上可以获得较好的实验结果。提出的增强特征融合网络能提取辨别性较强的行人特征,该模型能够应用于大场景非重叠多摄像机下的行人再识别问题,具有较高的识别能力和识别精度,且对背景变化的行人图像能提取具有较强的鲁棒性特征。     

PMA-Net: A parallelly mixed attention network for person re-identification

The objective of person re-identification (Re-ID) is to match unmistakable people across different settings and camera views. Although the use of convolutional neural networks (CNN) has been effective, they lose much significant data and ignore the correlation between the overall and partial features brought about by downsampling operators (e.g., pooling). This compromises the retrieval performance based on the Euclidean distance. We propose a parallelly mixed attention network (PMA-Net) to demonstrate the complementarity of spatial and channel information for Re-ID. First, we integrate self-attention with depthwise convolution in a parallel design, extracting different dimensions of information. Second, we propose two interactive branches to entwine spatial and channel information while extracting features. Then, we add a new additional embedding to incorporate non-visual inputs and reduce the feature bias toward camera variances. Finally, we perform additional maximizing the minimal pairwise angles (MMA) regularization on the features of PMA, which successfully encourages the angular diversity of feature vectors. We confirm the effectiveness of PMA-Net through an extensive set of ablation studies. The results of three benchmarks show the superiority of our model over current mAP/top1 accuracies of 91.3%/96.1% on Market-1501, 83.3%/91.7% on DukeMTMC-ReID, and 66.2%/85.1% on MSMT17.     

Relation network based on multi-granular hypergraphs for person re-identification

Applied Intelligence - In recent years, person re-identification (re-ID) has become a widespread research topic that focuses on retrieving target pedestrians from a set of images, typically taken...     

Cross-modality person re-identification via channel-based partition network

Applied Intelligence - Visible-infrared cross-modality person re-identification is an important task in the night video surveillance system, the huge difference between infrared and visible light...     

Gait recognition for person re-identification

The Journal of Supercomputing - Person re-identification across multiple cameras is an essential task in computer vision applications, particularly tracking the same person in different scenes....     

基于联合损失胶囊网络的换衣行人重识别

目前的行人重识别（Re-ID）研究主要集中在短时间情形,即一个人的衣着不太可能发生改变的情况。然而现实中更常见的是长时间的情况,这时一个人有很大的机会更换衣服,Re-ID模型应该考虑这种情况。为此,研究了一种基于联合损失胶囊网络的换衣行人重识别方法。所提方法基于换衣行人重识别胶囊网络ReIDCaps,使用与传统的标量神经元相比包含更多信息的矢量胶囊,用其长度表示行人身份信息,用其方向表示行人衣着信息;采用软嵌入注意力（SEA）防止模型过拟合;使用特征稀疏表示（FSR）机制提取具有判别性的特征;增加标签平滑正则化交叉熵损失与Circle Loss的联合损失以提高模型的泛化能力和鲁棒性。在三个换衣行人重识别数据集Celeb-reID、Celeb-reID-light和NKUP上进行实验,实验结果表明所提方法与目前已有的Re-ID方法相比具有一定优势。     

Instance interactive association graph convolutional network for domain adaptive person re-identification

Applied Intelligence - Domain adaptive person re-identification (re-ID) is a challenging task due to the large domain divergency between different datasets and the complicated variations of the...     

Occluded person re-identification based on differential attention siamese network

Applied Intelligence - Occluded person re-identification (ReID) aims to retrieve images of the same person with occlusion. The key is excavating discriminative features from the nonoccluded regions...     

A benchmark for clothes variation in person re-identification

Person re-identification (re-ID) has drawn attention significantly in the computer vision society due to its application and research significance. It aims to retrieve a person of interest across different camera views. However, there are still several factors that hinder the applications of person re-ID. In fact, most common data sets either assume that pedestrians do not change their clothing across different camera views or are taken under constrained environments. Those constraints simplify the person re-ID task and contribute to early development of person re-ID, yet a person has a great possibility to change clothes in real life. To facilitate the research toward conquering those issues, this paper mainly introduces a new benchmark data set for person re-identification. To the best of our knowledge, this data set is currently the most diverse for person re-identification. It contains 107 persons with 9,738 images, captured in 15 indoor/outdoor scenes from September 2019 to December 2019, varying according to viewpoints, lighting, resolutions, human pose, seasons, backgrounds, and clothes especially. We hope that this benchmark data set will encourage further research on person re-identification with clothes variation. Moreover, we also perform extensive analyses on this data set using several state-of-the-art methods. Our dataset is available at https://github.com/nkicsl/NKUP-dataset .     

基于孪生网络和双向最大边界排序损失的行人再识别

针对在实际场景中存在的不同行人图像之间比相同行人图像之间更相似所造成的行人再识别准确率较低的问题，提出一种基于孪生网络并结合识别损失和双向最大边界排序损失的行人再识别方法。首先，对在超大数据集上预训练过的神经网络模型进行结构改造，主要是对最后的全连接层进行改造，使模型可以在行人再识别数据集上进行识别判断；其次，联合识别损失和排序损失监督网络在训练集上的训练，并通过正样本对的相似度值减去负样本对的相似度值大于预定阈值这一判定条件，来使得负例图像对之间的距离大于正例图像对之间的距离；最后，使用训练好的神经网络模型在测试集上测试，提取特征并比对特征之间的余弦相似度。在公开数据集Market-1501、CUHK03和DukeMTMC-reID上进行的实验结果表明，所提方法分别取得了89.4%、86.7%、77.2%的rank-1识别率，高于其他典型的行人再识别方法，并且该方法在基准网络结构下最高达到了10.04%的rank-1识别率提升。     

Robust joint learning network: improved deep representation learning for person re-identification

Multimedia Tools and Applications - Existing person re-identification methods, which based on deep representation learning, mostly only focus on either global feature or local feature. This...