WO2020001083A1 - 一种基于特征复用的人脸识别方法 - Google Patents

一种基于特征复用的人脸识别方法 Download PDF

Info

Publication number
WO2020001083A1
WO2020001083A1 PCT/CN2019/078473 CN2019078473W WO2020001083A1 WO 2020001083 A1 WO2020001083 A1 WO 2020001083A1 CN 2019078473 W CN2019078473 W CN 2019078473W WO 2020001083 A1 WO2020001083 A1 WO 2020001083A1
Authority
WO
WIPO (PCT)
Prior art keywords
feature
sample
tested
data set
identity
Prior art date
Application number
PCT/CN2019/078473
Other languages
English (en)
French (fr)
Inventor
陆生礼
庞伟
周世豪
向家淇
范雪梅
泮雯雯
Original Assignee
东南大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 东南大学 filed Critical 东南大学
Publication of WO2020001083A1 publication Critical patent/WO2020001083A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/168—Feature extraction; Face representation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/172—Classification, e.g. identification

Definitions

  • the invention discloses a method for face recognition based on feature multiplexing, belonging to the technical field of computational estimation, and particularly to the field of computer vision technology for face recognition.
  • Face recognition technology has been widely used in access control, security inspection, monitoring, etc. Its main task is to distinguish different individuals in the database and reject individuals outside the database. In practical applications, the features of a person's appearance will be affected by dressing, expressions, and changes due to posture, lighting, and the front picture of the same person will also change over time. In order to increase the robustness of the algorithm, it is necessary to update the model in specific situations during the recognition process. The traditional method is to re-collect the samples and train again, which is time-consuming and difficult to operate.
  • Existing online learning methods compare the shallow features of a face (such as Haar features, LBP features) to identify and track a given face in a video.
  • the target face is distinguished from one or more surrounding faces, and only a few samples need to be discerned; at the same time, the facial features change little during the short period of time included in the video, so the image
  • the shallow features can represent the facial features to a certain extent.
  • tasks such as face access control and time attendance need to distinguish a database containing hundreds of people. Over a long period of time, everyone's appearance will change, and shallow features are difficult to handle such complex tasks.
  • Deep neural network improves the recognition of the model, but the training of the network consumes a lot of computing resources and time.
  • the model trained on the offline server needs to be re-imported into the face recognition device.
  • the neural network structure is fixed , When adding / deleting members, it is necessary to train again, which brings inconvenience to practical applications.
  • the above-mentioned face recognition technology has the disadvantages of large calculation amount, occupying more computing resources, and the accuracy rate needs to be improved.
  • this application aims to propose a feature-based human reuse Face recognition method.
  • the object of the present invention is to address the shortcomings of the background art described above, and provide a method for face recognition based on feature multiplexing, which can quickly and accurately identify faces with limited computing resources, and solves the complicated calculation and occupation of existing face recognition technologies. More computing resources and technical issues that need to be improved in accuracy.
  • Establish external data sets Establish external data sets based on public face databases of research institutions or self-collected data.
  • the face databases can choose public databases such as CASIA-WebFace, VGG-FACE; or they can be captured on the network by themselves Take pictures of public figures.
  • Each picture should include an identification that indicates which individual the picture belongs to. You should collect as many individuals as possible, each individual containing as many samples as possible, while reducing the number of mislabeled samples in the data set. Increasing the number of samples and the number of categories will improve the training accuracy, and will not change the structure of the face feature extractor or increase the training difficulty;
  • a local member set U ⁇ u 1 , u 2 , ..., u m ⁇ composed of m individuals, and take n corresponding face samples ⁇ x i1 for each member u i in U , x i2 , ..., x in ⁇ , preferably, the face samples should be photos with normal lighting and natural expressions.
  • Training model Use a convolutional neural network as a feature extractor.
  • the input of the neural network is a color picture.
  • the output of the neural network is the category to which the picture belongs.
  • the length of the classification layer is equal to the number of categories in the external data set.
  • the loss function can use softmaxloss. The reason is that the neural network is trained with an external data set, because the number and type of samples in the external data set are much larger than the local data set, which is conducive to the neural network to learn better features.
  • the loss function is continuously reduced with the back propagation of the error. The accuracy rate keeps increasing. When the loss function converges and no longer decreases, save the convolutional neural network model and use the l-dimensional vector connected to the classification layer as the feature vector of the input picture.
  • the convolutional neural network involved in this application adds at least one dense connection block for hierarchically extracting features in the network.
  • Each dense connection block is responsible for extracting first-level features.
  • Each dense connection block contains at least two convolutions connected in sequence. Layer, the feature map output by the current convolution layer and the feature map output by all convolution layers before the convolution layer are stitched as the input feature map to the next convolution layer. The feature map output by each densely connected block is reduced. After sampling, it is transmitted to the input end of the next dense connection block;
  • 2 , d represents the similarity between the two features. The larger d, the larger the feature gap, and further, when d is large enough, the two features can be considered to belong to different individuals, and the reference vector closest to y in S is found. And distance i * argmin i ⁇ ⁇ 1,2, ..., m ⁇ d i , set the similarity threshold ⁇ , if Output Otherwise output u represents the identity of the person predicted by the model.
  • the color face image of the input convolutional neural network is processed by a plurality of convolutional layers and downsampling layers of equal steps to obtain a feature map of the first densely connected block, and the output of the last densely connected block is The feature map is then subjected to convolution operations and mean pooling operations to obtain the feature vectors input to the classification layer.
  • this application also provides a face recognition method without retraining the model after adding / removing members.
  • new members provide their own true identity tags u k after completing the face recognition process
  • the present application also provides a terminal device for implementing the above-mentioned face recognition method.
  • the device includes: a memory, a processor, and a computer program stored on the memory and run on the processor.
  • the processor executes the program, the following steps are implemented: : Use external data sets to train face feature extractors, and extract the reference features corresponding to each member in the local data set hierarchically by multiple equal-step convolution and feature map stitching to form a reference feature space, compare the feature vectors of the samples to be tested and The reference feature is used to determine the reference feature most similar to the feature vector of the sample to be tested.
  • the member to which the reference feature most similar to the feature vector of the sample to be tested belongs belongs.
  • the identity of is the identity of the sample to be tested; otherwise, a message indicating that the identity of the sample to be tested fails is returned.
  • the present invention proposes a face recognition method with multiplexed features.
  • Feature extraction is achieved through densely connected convolutional neural networks.
  • a densely connected layer is formed by connecting several synchronizing convolutional layers, and the output features of each convolutional layer
  • the map and all the output feature maps of the previous convolutional layer are stitched to make the input feature map of the next convolutional layer, which strengthens feature reuse, improves network performance, reduces the number of parameters and calculations, and is more robust and applicable.
  • the scope is wider, and the recognition speed and accuracy can be improved as much as possible with limited computing resources.
  • the feature-recognized face recognition method can also be extended to image recognition fields such as vehicle recognition and pedestrian recognition.
  • This application also provides a method for dynamically adding or deleting members at the terminal.
  • This method enables offline updating of the face recognition model by flexibly adjusting the reference feature space extracted from the local dataset to adapt to changes in the dataset. Compared with the traditional method of re-collecting samples and training again, the operation is simple and the amount of calculation is small.
  • the model does not need to be updated online when the data set is changed, and it is especially suitable for face recognition in offline occasions.
  • FIG. 1 is a flowchart of face recognition by this method.
  • Figure 2 is an example of a face cut sample from a data set.
  • FIG. 3 is a schematic structural diagram of a dense connection block.
  • FIG. 1 shows a flowchart of a face recognition method according to the present invention.
  • the face recognition method includes the following five steps.
  • Step 1 Establish an external data set:
  • the CASIA-WebFace database is used as the external data set.
  • Figure 2 shows a sample sample of the processed CASIA-WebFace database.
  • the face frame should fit the person more closely.
  • all pictures are scaled to the input size of the convolutional neural network. If external data sets are obtained from other data sets, it is also necessary to follow the processing method in which the face frame closely fits the edge of the face and the picture meets the input picture size requirements of the neural network.
  • Step 2 Establish a local data set: take pictures of the faces of ten people, and take multiple face sample pictures of each person with different expressions and poses.
  • Step 3 Establish a convolutional neural network: use the external data set as a sample set to train a facial feature extractor: This application relates to a more efficient convolutional neural network.
  • the input of the neural network is 160 * A 160-pixel color face picture.
  • the color face picture first passes through three convolution layers with a step size of 1 and a down-sampling layer to obtain a feature map of 80 * 80.
  • the feature map of 80 * 80 is then input to the first
  • the dense connection block is used as the input feature map of the first dense connection block.
  • the dense connection block contains three convolutional layers.
  • the input feature map is first input to convolutional layer 1.
  • the input feature map is concatenated with the output feature map of convolutional layer 1 and input to convolutional layer 2.
  • the convolutional layer 1 and convolutional layer 2 The output feature map is spliced and input to the convolution layer 3. Downsample the output feature map of convolution layer 3 to 40 * 40 and input the next dense connection block, repeat the same operation. After three densely connected blocks, the size of the feature map becomes 20 * 20, and the feature map of 20 * 20 then passes through two convolution layers with a step size of 2 to obtain 64 3 * 3 feature maps and 64 3 * 3 feature maps.
  • Feature map input mean pooling layer to obtain 64-dimensional feature vectors.
  • the category of the training picture is output at the classification layer, and the error is calculated and back-propagated.
  • the features of the picture to be tested are output at the feature layer, and the neural network is trained until the loss function converges. Is h (x).
  • 2 to find the reference feature vector closest to y in S And distance i * argmin i ⁇ ⁇ 1,2, ..., m ⁇ d i , set the similarity threshold ⁇ , if Output Otherwise, the output A larger ⁇ represents a more relaxed judgment criterion, and a more relaxed judgment criterion tends to regard the test subject as a member of the local data set; a smaller ⁇ does the opposite.
  • the face recognition method provided in this application may be implemented on a terminal device, the device including at least one memory including a update member button, a delete member button, an input module, a computer software program storing the above-mentioned face recognition method, and a processor.
  • the input module may be a card swiping device or a keyboard for a test subject to input his or her identity tag.
  • the system suspends video streaming and saves the current input picture x and the prediction result.
  • the device may further include a permission obtaining module.
  • the invention also provides a simple way to add / remove members.
  • the system suspends video streaming and removes the information of the member to be deleted from the local individual set U and the reference feature space S.
  • the administrator is given permission to add / remove members through the device's Get Permissions module.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Collating Specific Patterns (AREA)
  • Image Analysis (AREA)

Abstract

一种基于特征复用的人脸识别方法，属于计算推算的技术领域，尤其涉及人脸识别的计算机视觉技术领域。该方法利用外部数据集训练人脸特征提取器，通过多次等步长卷积及特征图拼接的方式分级提取本地数据集中各成员对应的参考特征以构成参考特征空间，对比待测试样本的特征向量和参考特征以确定与待测试样本的特征向量最相似的参考特征，在与待测试样本的特征向量最相似的参考特征满足阈值要求时，以与待测试样本的特征向量最相似的参考特征所属成员的身份为待测试样本的身份，否则，返回待测试样本身份识别失败的消息，以较少的计算资源实现了人脸的快速识别。

Description

一种基于特征复用的人脸识别方法 技术领域
本发明公开了一种基于特征复用的人脸识别方法，属于计算推算的技术领域，尤其涉及人脸识别的计算机视觉技术领域。
背景技术
人脸识别技术已经广泛运用于门禁、安检、监控等方面，其主要任务是区分数据库中的不同个体并拒绝数据库之外的个体。在实际应用中，人的相貌特征会受到装扮、表情的影响且因姿势、光照而变化，同一个人的正面图片也会随时间的推移而出现不同。为增加算法的鲁棒性，在识别过程中，有必要在特定情况下更新模型。传统的方法是重新收集样本再次训练，这种做法费时费力，难以操作。
现有的在线学习方法通过提取人脸的浅层特征(如：Haar特征、LBP特征)进行比对，在视频中识别并跟踪给定的人脸。在这种应用场景下，把目标人脸和周围的一个或多个人脸区分开，只需要辨别很少的样本；同时，在视频包含的小段时间内，人脸特征变化较小，因此，图像的浅层特征可以在一定程度上表征人脸特征。但是，人脸门禁、考勤等任务需要分辨包含数百人的数据库，在相当长的一段时间内，每个人的相貌都会有所改变，浅层特征难以处理如此复杂的任务。
深度神经网络提高了模型的辨识度，但网络的训练耗费大量的运算资源和时间，变更模型时需要将在离线服务器上训练好的模型重新导入人脸识别设备；另一方面，神经网络结构固定，增加/删除成员时同样需要再次训练，为实际应用带来不便。
上述人脸识别技术存在计算量大、占用较多计算资源、准确率有待提高的缺陷，为提高人脸识别准确率并减少占用的计算机资源，本申请旨在提出一种基于特征复用的人脸识别方法。
发明内容
本发明的发明目的是针对上述背景技术的不足，提供了一种基于特征复用的人脸识别方法，以有限的计算资源快速准确地识别人脸，解决了现有人脸识别技术计算复杂、占用较多计算资源、准确率有待提高的技术问题。
本发明为实现上述发明目的采用如下技术方案：
一种基于特征复用的人脸识别方法，
建立外部数据集：根据研究机构的公开人脸数据库或自行搜集的数据建立外部数据集，示例性地，人脸数据库可以选择CASIA-WebFace、VGG-FACE等公开数据库；也可以自行在网络上抓取公众人物的图片。每张图片都应含有身份标注，指明该图片属于哪个个体。应当收集尽可能多的个体，每个个体包含尽可能多的样本，同时减少数据集中错误标注样本的数量。样本数量和类别数量的增加会提高训练精度，且不会改变人脸特征提取器的结构或增加训练难度；
建立本地数据集：假设由m个人组成本地成员集合U＝{u 1,u 2,...,u m}，给U中的每个成员u i拍摄n张对应的人脸样本{x i1,x i2,...,x in}，优选地，人脸样本应该是光照正常、表情自然的照片，当条件允许拍摄多张图片时，可以关注表情和姿态的多样性；
训练模型：使用卷积神经网络作为特征提取器，神经网络的输入为彩色图片，神经网络的输出为图片所属类别，分类层的长度等于外部数据集的类别数，损失函数可以采用softmaxloss，需要说明的是，神经网络采用外部数据集训练，因为外部数据集的样本数量和种类远超本地数据集，有利于神经网络学习到更好的特征，损失函数随着误差的反向传播不断下降，训练准确率不断上升，当损失函数收敛并不再继续下降时，保存卷积神经网络模型，把与分类层相连的l维向量作为输入图片的特征向量，特征向量的维度一般远小于类别数量，可以取几十到几百之间，记输入图片x到特征向量的映射为h(x)，用训练好的特征提取器提取本地数据集的样本特征，计算得到每个个体对应的参考特征
Figure PCTCN2019078473-appb-000001
其中，n代表人脸库中第i个人的人脸样本个数，建立参考特征空间S＝{y 1,y 2,...,y m}，
本申请涉及的卷积神经网络在网络中增加了至少一个用于分级提取特征的稠密连接块，每个稠密连接块负责提取一级特征，每个稠密连接块至少包含两个依次连接的卷积层，当前卷积层输出的特征图和该卷积层之前所有卷积层输出的特征图拼接后作为至下一卷积层的输入特征图，每一个稠密连接块输出的特征图都经降采样后传输至下一稠密连接块的输入端；
预测待测图片所属个体的身份：在视频帧中截取待测者的人脸区域图片，处理所截图片得到待测图片x，使用特征提取器提取待测图片x的特征向量y，y＝h(x)，对所有y i∈S计算y与y i的距离d：d i＝||y-y i|| 2，d表征了两个特征之间的相似度。d越大表征特征差距就越大，更进一步地，当d足够大时，可以认为两个特征属于不同的个体，找出S中与y距离最近的参考向量
Figure PCTCN2019078473-appb-000002
以及距离
Figure PCTCN2019078473-appb-000003
i *＝argmin i∈{1,2,...,m}d i，设定相似度阈值δ，如果
Figure PCTCN2019078473-appb-000004
输出
Figure PCTCN2019078473-appb-000005
否则输出
Figure PCTCN2019078473-appb-000006
u代表模型预测的待测者身份。
优选地，输入卷积神经网络的彩色人脸图片经多个等步长的卷积层和降采样层的处理后得到输入第一个稠密连接块的特征图，对最后一个稠密连接块输出的特征图再进行卷积操作和均值池化操作得到输入至分类层的特征向量。
进一步的，本申请还提供了一种添加/删除成员后无需重新训练模型的人脸识别方法，添加成员时，新成员完成一次人脸识别过程后提供自己的真实身份标签u k，
Figure PCTCN2019078473-appb-000007
暂停视频流传输，保存当前输入图片x及特征提取器从当前图片提取的特征向量y，更新本地成员集合为U′，U′＝U∪u k，更新参考特征空间为S′，S′＝S∪y，更新完毕后恢复视频流；删除成员时，暂停视频流传输，在本地成员集合U和参考特征空间S中移除待删除成员的信息，恢复视频流。
本申请还提供了一种实现上述人脸识别方法的终端设备，该设备包括：存储器、处理器及存储在存储器上并在处理器上运行的计算机程序，处理器执行所述程序时实现以下步骤：利用外部数据集训练人脸特征提取器，通过多次等步长卷积及特征图拼接的方式分级提取本地数据集中各成员对应的参考特征以构成参考特征空间，对比待测试样本的特征向量和参考特征以确定与待测试样本的特征向量最相似的参考特征，在与待测试样本的特征向量最相似的参考特征满足阈值要求时，以与待测试样本的特征向量最相似的参考特征所属成员的身份为待测试样本的身份，否则，返回待测试样本身份识别失败的消息。
本发明采用上述技术方案，具有以下有益效果：
(1)本发明提出了复用特征的人脸识别方法，通过稠密连接的卷积神经网络实现特征提取，通过连接若干同步长的卷积层构成稠密连接层，每个卷积层的输出特征图和之前卷积层的所有输出特征图拼接后作出下一卷积层的输入特征图，强化了特征复用，提升了网络性能，减少了参数数量和运算量，鲁棒性更强，适用范围更广，以有限的计算资源尽可能提高识别速度和准确率，该特征复用的人脸识别方法还可以推广至车辆识别、行人识别等图像识别领域。
(2)本申请还提供了一种在终端动态添加或删除成员的方法，该方法通过灵活调整从本地数据集提取的参考特征空间以适应数据集的变化，实现了人脸识别模型的离线更新，相比于重新收集样本再次训练的传统方法，操作简单，计算量小，在数据集发成变化时无需对模型进行在线更新，尤其适合应用于离线场合的人脸识别。
附图说明
图1是本方法的人脸识别流程图。
图2是数据集的人脸截取样例。
图3是稠密连接块的结构示意图。
具体实施方式
为了更清楚地说明本发明的特征，下面结合附图和具体实施方式进行进一步的详细描述。需要说明的是，以下的阐述提到了许多具体细节以便于充分理解本发明，本发明包括但不限于以下实施方例。
图1给出了根据本发明人脸识别方法的流程图，该人脸识别方法包括以下五个步骤。
步骤一、建立外部数据集：采用CASIA-WebFace数据库作为外部数据集，图2给出了经过处理的CASIA-WebFace数据库的样本实例，如图2所示，人脸框应该比较紧密地贴合人脸边缘，所有图片缩放成卷积神经网络的输入尺寸。如从其它数据集获取外部数据集，也需要遵循人脸框紧密地贴合人脸边缘以及图片满足神经网络输入图片尺寸要求的处理方式。
步骤二、建立本地数据集：拍摄十个人的脸部照片，拍摄每一个人表情和姿态不同的多张人脸样本图片。
步骤三、建立卷积神经网络：以外部数据集合为样本集训练人脸特征提取器：本申请涉及了一种更高效的卷积神经网络，如图3所示，神经网络的输入是160*160像素的彩色人脸图片，彩色人脸图片首先依次经过三个步长为1的卷积层和一个降采样层得到80*80的特征图，80*80的特征图随后输入至第一个稠密连接块做为第一个稠密连接块的输入特征图。稠密连接块包含三个卷积层，输入特征图首先输入卷积层1，输入特征图与卷积层1的输出特征图拼接后输入卷积层2；卷积层1和卷积层2的输出特征图拼接后输入卷积层3。将卷积层3的输出特征图降采样到40*40后输入下一个稠密连接块，重复相同的操作。经过三个稠密连接块后，特征图大小变为20*20，20*20的特征图随后经过两次步长为2的卷积层得到64个3*3的特征图，64个3*3特征图输入均值池化层得到64维特征向量。训练时，在分类层输出训练图片所属类别，计算误差并反向传播；测试时，在特征层输出待测图片的特征，训练神经网络直到损失函数收敛，记此时神经网络输入到输出的映射为h(x)。
步骤四、构建参考特征空间：由训练后的人脸特征提取器提取本地样本集的特征，计算得到每个个体对应的参考特征y i，
Figure PCTCN2019078473-appb-000008
本地样本集中各个体对应的参考特征构成参考特征空间S，S＝{y 1,y 2,...,y m}。
步骤五、对比待测样本的预测特征向量和参考特征空间中的各参考特征向量确定待测样本所属个体：使用训练好的特征提取器预测待测图片x的特征向量y，y＝h(x)，对所有y i∈S，计算y与y i的距离：d i＝||y-y i|| 2，找出S中与y距离最近的参考特征向量
Figure PCTCN2019078473-appb-000009
以及距离
Figure PCTCN2019078473-appb-000010
i *＝argmin i∈{1,2,...,m}d i，设定相似度阈值δ，如果
Figure PCTCN2019078473-appb-000011
输出
Figure PCTCN2019078473-appb-000012
否则，输出
Figure PCTCN2019078473-appb-000013
较大的δ代表更宽松的判断标准，宽松的判断标准更倾向于把待测者看作本地数据集的某个成员；较小的δ反之。
本申请提供的人脸识别方法可在终端设备上实现，该设备包括至少一个包含更新成员按键、一个删除成员按键、一个输入模块、存储有上述人脸识别方法的计算机软件程序的存储器及处理器。示例性的，输入模块可以是供待测者输入自己的身份标签的刷卡装置或键盘。系统暂停视频流传输，保存当次的输入图片x和预测结果。可选地，设备还可以包括获取权限模块。
本发明还提供了一种简便的添加/删除成员方式。添加成员时，新成员完成一次人脸识别过程，通过设备的输入模块提供自己的真实身份标签，发出添加成员指令(待测者按下更新成员按键)后，系统暂停视频流传输，保存当次的输入图片x和特征向量y，更新本地个体集合U′＝U∪u k，更新参考特征空间S′＝S∪y k，y k＝y；删除成员时，待测试者通过输入模块提供待删除的成员标签，发出删除成员(待测试者按下删除成员按键)指令后，系统暂停视频流传输，在本地个体集合U和参考特征空间S中移除待删除成员的信息。通过设备的获取权限模块授予管理员添加/删除成员的权限。

Claims (10)

  1. 一种基于特征复用的人脸识别方法，其特征在于，利用外部数据集训练人脸特征提取器，通过多次等步长卷积及特征图拼接的方式分级提取本地数据集中各成员对应的参考特征以构成参考特征空间，对比待测试样本的特征向量和参考特征以确定与待测试样本的特征向量最相似的参考特征，在与待测试样本的特征向量最相似的参考特征满足阈值要求时，以与待测试样本的特征向量最相似的参考特征所属成员的身份为待测试样本的身份，否则，返回待测试样本身份识别失败的消息。
  2. 根据权利要求1所述一种基于特征复用的人脸识别方法，其特征在于，人脸特征提取器通过包含至少一个稠密连接块的卷积神经网络实现，每个稠密连接块包含至少两个依次连接的同步长卷积层，当前卷积层输出的特征图和该卷积层之前所有卷积层输出的特征图拼接后作为至下一卷积层的输入特征图，每一个稠密连接块输出的特征图都经降采样后传输至下一稠密连接块的输入端。
  3. 根据权利要求2所述一种基于特征复用的人脸识别方法，其特征在于，对最后一个稠密连接块输出的特征图再进行卷积操作和均值池化操作得到输入至分类层的特征向量。
  4. 根据权利要求2所述一种基于特征复用的人脸识别方法，其特征在于，输入至第一个稠密连接块的特征图通过对输入网络的初始样本进行卷积及降采样操作获取。
  5. 根据权利要求1所述一种基于特征复用的人脸识别方法，其特征在于，所述外部数据集从公开数据库中选择样本或自行在网络上抓取的人物图片。
  6. 根据权利要求1所述一种基于特征复用的人脸识别方法，其特征在于，所述本地数据集包含本地成员集合及各本地成员对应的人脸样本构成的人脸集合。
  7. 根据权利要求1所述一种基于特征复用的人脸识别方法，其特征在于， 添加本地成员时，将新添加成员的身份信息添加至本地数据集，提取新添加成员图片的特征并将所提取的特征添加至参考特征空间。
  8. 根据权利要求1所述一种基于特征复用的人脸识别方法，其特征在于，删除成员时，从本地数据集及参考特征空间移除待删除成员的数据。
  9. 一种计算机可读存储介质，其上存储有计算机程序，其特征在于，该程序被处理器执行时实现权利要求1所述的方法。
  10. 一种人脸识别终端设备，包括：存储器、处理器及存储在存储器上并在处理器上运行的计算机程序，其特征在于，所述处理器执行所述程序时实现以下步骤：利用外部数据集训练人脸特征提取器，通过多次等步长卷积及特征图拼接的方式分级提取本地数据集中各成员对应的参考特征以构成参考特征空间，对比待测试样本的特征向量和参考特征以确定与待测试样本的特征向量最相似的参考特征，在与待测试样本的特征向量最相似的参考特征满足阈值要求时，以与待测试样本的特征向量最相似的参考特征所属成员的身份为待测试样本的身份，否则，返回待测试样本身份识别失败的消息。
PCT/CN2019/078473 2018-06-30 2019-03-18 一种基于特征复用的人脸识别方法 WO2020001083A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810702467.9A CN109214263A (zh) 2018-06-30 2018-06-30 一种基于特征复用的人脸识别方法
CN201810702467.9 2018-06-30

Publications (1)

Publication Number Publication Date
WO2020001083A1 true WO2020001083A1 (zh) 2020-01-02

Family

ID=64989797

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/078473 WO2020001083A1 (zh) 2018-06-30 2019-03-18 一种基于特征复用的人脸识别方法

Country Status (2)

Country Link
CN (1) CN109214263A (zh)
WO (1) WO2020001083A1 (zh)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111274886A (zh) * 2020-01-13 2020-06-12 天地伟业技术有限公司 一种基于深度学习的行人闯红灯违法行为分析方法及系统
CN111339990A (zh) * 2020-03-13 2020-06-26 乐鑫信息科技（上海）股份有限公司 一种基于人脸特征动态更新的人脸识别系统和方法
CN111814702A (zh) * 2020-07-13 2020-10-23 安徽兰臣信息科技有限公司 一种基于成年人脸和儿童照特征空间映射关系的儿童人脸识别方法
CN112183449A (zh) * 2020-10-15 2021-01-05 上海汽车集团股份有限公司 一种驾驶员身份验证方法、装置、电子设备及存储介质
CN112329890A (zh) * 2020-11-27 2021-02-05 北京市商汤科技开发有限公司 图像处理方法及装置、电子设备及存储介质
CN113723247A (zh) * 2021-08-20 2021-11-30 西安交通大学 一种脑电身份识别方法及系统
CN113989886A (zh) * 2021-10-22 2022-01-28 中远海运科技股份有限公司 基于人脸识别的船员身份验证方法
CN114613058A (zh) * 2022-03-25 2022-06-10 中国农业银行股份有限公司 一种具有考勤功能的门禁系统、考勤方法和相关装置

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109214263A (zh) * 2018-06-30 2019-01-15 东南大学 一种基于特征复用的人脸识别方法
CN110378092B (zh) * 2019-07-26 2020-12-04 北京积加科技有限公司 身份识别系统及客户端、服务器和方法
CN111414941B (zh) * 2020-03-05 2023-04-07 清华大学深圳国际研究生院 一种基于特征复用的点云卷积神经网络

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102982321A (zh) * 2012-12-05 2013-03-20 深圳Tcl新技术有限公司 人脸库采集方法及装置
CN106650694A (zh) * 2016-12-30 2017-05-10 江苏四点灵机器人有限公司 一种以卷积神经网络作为特征提取器的人脸识别方法
CN107133579A (zh) * 2017-04-20 2017-09-05 江南大学 基于CSGF(2D)2PCANet卷积网络的人脸识别方法
CN107679531A (zh) * 2017-06-23 2018-02-09 平安科技（深圳）有限公司 基于深度学习的车牌识别方法、装置、设备及存储介质
CN109214263A (zh) * 2018-06-30 2019-01-15 东南大学 一种基于特征复用的人脸识别方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102982321A (zh) * 2012-12-05 2013-03-20 深圳Tcl新技术有限公司 人脸库采集方法及装置
CN106650694A (zh) * 2016-12-30 2017-05-10 江苏四点灵机器人有限公司 一种以卷积神经网络作为特征提取器的人脸识别方法
CN107133579A (zh) * 2017-04-20 2017-09-05 江南大学 基于CSGF(2D)2PCANet卷积网络的人脸识别方法
CN107679531A (zh) * 2017-06-23 2018-02-09 平安科技（深圳）有限公司 基于深度学习的车牌识别方法、装置、设备及存储介质
CN109214263A (zh) * 2018-06-30 2019-01-15 东南大学 一种基于特征复用的人脸识别方法

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111274886A (zh) * 2020-01-13 2020-06-12 天地伟业技术有限公司 一种基于深度学习的行人闯红灯违法行为分析方法及系统
CN111274886B (zh) * 2020-01-13 2023-09-19 天地伟业技术有限公司 一种基于深度学习的行人闯红灯违法行为分析方法及系统
CN111339990B (zh) * 2020-03-13 2023-03-24 乐鑫信息科技（上海）股份有限公司 一种基于人脸特征动态更新的人脸识别系统和方法
CN111339990A (zh) * 2020-03-13 2020-06-26 乐鑫信息科技（上海）股份有限公司 一种基于人脸特征动态更新的人脸识别系统和方法
CN111814702A (zh) * 2020-07-13 2020-10-23 安徽兰臣信息科技有限公司 一种基于成年人脸和儿童照特征空间映射关系的儿童人脸识别方法
CN112183449A (zh) * 2020-10-15 2021-01-05 上海汽车集团股份有限公司 一种驾驶员身份验证方法、装置、电子设备及存储介质
CN112183449B (zh) * 2020-10-15 2024-03-19 上海汽车集团股份有限公司 一种驾驶员身份验证方法、装置、电子设备及存储介质
CN112329890A (zh) * 2020-11-27 2021-02-05 北京市商汤科技开发有限公司 图像处理方法及装置、电子设备及存储介质
CN113723247A (zh) * 2021-08-20 2021-11-30 西安交通大学 一种脑电身份识别方法及系统
CN113723247B (zh) * 2021-08-20 2024-04-02 西安交通大学 一种脑电身份识别方法及系统
CN113989886A (zh) * 2021-10-22 2022-01-28 中远海运科技股份有限公司 基于人脸识别的船员身份验证方法
CN113989886B (zh) * 2021-10-22 2024-04-30 中远海运科技股份有限公司 基于人脸识别的船员身份验证方法
CN114613058A (zh) * 2022-03-25 2022-06-10 中国农业银行股份有限公司 一种具有考勤功能的门禁系统、考勤方法和相关装置

Also Published As

Publication number Publication date
CN109214263A (zh) 2019-01-15

Similar Documents

Publication Publication Date Title
WO2020001083A1 (zh) 一种基于特征复用的人脸识别方法
WO2020001084A1 (zh) 一种在线学习的人脸识别方法
WO2021227726A1 (zh) 面部检测、图像检测神经网络训练方法、装置和设备
CN110135249B (zh) 基于时间注意力机制和lstm的人体行为识别方法
CN110555481A (zh) 一种人像风格识别方法、装置和计算机可读存储介质
WO2021218238A1 (zh) 图像处理方法和图像处理装置
CN109063626B (zh) 动态人脸识别方法和装置
WO2021063056A1 (zh) 人脸属性识别方法、装置、电子设备和存储介质
CN111444826A (zh) 视频检测方法、装置、存储介质及计算机设备
Fang et al. Dynamic gesture recognition using inertial sensors-based data gloves
CN110175515B (zh) 一种基于大数据的人脸识别算法
US11631277B2 (en) Change-aware person identification
CN113780249B (zh) 表情识别模型的处理方法、装置、设备、介质和程序产品
CN113052150B (zh) 活体检测方法、装置、电子设备和计算机可读存储介质
CN110516533A (zh) 一种基于深度度量的行人再辨识方法
CN111242019A (zh) 视频内容的检测方法、装置、电子设备以及存储介质
Xia et al. Face occlusion detection using deep convolutional neural networks
CN113205002A (zh) 非受限视频监控的低清人脸识别方法、装置、设备及介质
Gupta et al. Single attribute and multi attribute facial gender and age estimation
CN112036284A (zh) 图像处理方法、装置、设备及存储介质
Echoukairi et al. Improved Methods for Automatic Facial Expression Recognition.
CN114519863A (zh) 人体重识别方法、人体重识别装置、计算机设备及介质
Abayomi-Alli et al. Facial image quality assessment using an ensemble of pre-trained deep learning models (EFQnet)
CN114663910A (zh) 基于多模态学习状态分析系统
CN113824989A (zh) 一种视频处理方法、装置和计算机可读存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19827462

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19827462

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 19827462

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS (EPO FORM 1205A DATED 10.08.2021)

122 Ep: pct application non-entry in european phase

Ref document number: 19827462

Country of ref document: EP

Kind code of ref document: A1