CN110427888A - A kind of face method for evaluating quality based on feature clustering - Google Patents
A kind of face method for evaluating quality based on feature clustering Download PDFInfo
- Publication number
- CN110427888A CN110427888A CN201910715538.3A CN201910715538A CN110427888A CN 110427888 A CN110427888 A CN 110427888A CN 201910715538 A CN201910715538 A CN 201910715538A CN 110427888 A CN110427888 A CN 110427888A
- Authority
- CN
- China
- Prior art keywords
- face
- data
- feature
- training
- quality
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The present invention relates to technical field of computer vision, and disclose a kind of face method for evaluating quality based on feature clustering, including the following steps: A, data preparation: training data of the present invention uses CMU Multi-PIE face database, it includes the different angle of different faces id that other, which can also be used, illumination, the human face photo of clarity；Multi-PIE data comprise more than the 750000+ face Various Seasonal of 337id, different angle, the human face photo of different illumination；B, it data prediction: before data are sent into training, needs to pre-process data, pretreated purpose is first is that in order to enrich sample number amount and type.The quality evaluation of face of the present invention is applied to the screening before face alignment, is clustered with face characteristic, and the data shape of face alignment needs is more in line with；Meanwhile this method is converted into trained label using the similarity of face characteristic sequence of calculation face characteristic, carries out neural metwork training, process is simple, and speed is fast, can satisfy requirement of real-time without being manually labeled.
Description
Technical field
The present invention relates to technical field of computer vision, specially a kind of face quality evaluation side based on feature clustering
Method.
Background technique
With the development of computer technology, the fields such as machine vision, neural network are had made great progress.Face among these
Identification technology achieves very ten-strike, has been widely used in bank, mobile phone, shopping, campus, subway, cell, public place
Etc. in each living scene.A set of face system includes face snap module, face quality assessment modules, face in monitoring scene
Comparison module and result output module, often due to scene is complicated, lead to the facial image matter captured in video monitoring scene
Measure it is irregular, since face obscures, blocks, causes the accuracy rate of recognition of face to be tested due to posture etc., thus in order to
The precision for improving identification generallys use a kind of method and judges the quality of facial image, the face that quality is met the requirements
System is given into matching identification, the false recognition rate of recognition of face can be effectively reduced in this way.
Existing face method for evaluating quality is roughly divided into two classes: one kind is based on conventional method, and one kind is based on depth
The method of study；Based on traditional method usually by the size of face, gradient such as blocks at the quality that conditions judge face: another
Outer one kind is the development with deep learning in recent years, and someone starts to apply deep learning in the assessment of face quality；
Hand-designed feature is needed based on traditional method, and is verified repeatedly for a large amount of data, is combined by various judgements
Achieve the purpose that face quality judging, the accuracy rate of the judgement of face quality is influenced by local module judgement；Based on depth
The face quality judging method of study needs a large amount of data mark, and the index of face quality annotation is numerous, between all multi objectives
Boundary is fuzzy, and artificial mark shows great difficulty.
Therefore, it is proposed that a kind of face method for evaluating quality based on feature clustering.
Summary of the invention
The present invention provides a kind of face method for evaluating quality based on feature clustering, solves and mentions in above-mentioned background technique
Out the problem of.
To achieve the above object, the invention provides the following technical scheme: a kind of face quality evaluation based on feature clustering
Method, including the following steps:
A, data preparation: training data of the present invention uses CMU Multi-PIE face database, it is possible to use other
Different angle comprising different faces id, illumination, the human face photo of clarity；Multi-PIE data comprise more than 337id's
750000+ face Various Seasonal, different angle, the human face photo of different illumination；
B, it data prediction: before data are sent into training, needs to pre-process data, pretreated purpose one
It is in order to enrich sample number amount and type, second is that in order to avoid over-fitting；
C, feature extraction: face characteristic extraction algorithm is using the face recognition to increase income, it is possible to use other people
Face knows method for distinguishing, and the certificate photo (everyone includes 3-5 of Various Seasonal) of 337id people in data set Multi-PIE uses
Face recognition algorithm extracts face characteristic and forms gallery, and gallery data are extracted feature and located in advance without image
Module is managed, image preprocessing is passed through for other data, by the image zooming-out face characteristic after pretreatment；
D, training label generates: it will currently be extracted face characteristic and the id certificate photo feature calculation similarity, it is similar
The distance between two feature of degree=1-, obtains the similarity between two faces, which is divided into 10 classes, as the face figure
The label 1 of picture, similarity value are the label 2 of the picture；
E, face quality model training: the data after pretreatment and two labels generated are fed together shown in Fig. 3
Depth network, in Fig. 3 after the last one convolution module 1, network is divided into Liang Ge branch, and a branch is used to return face
A convolution module 2 is passed through by score value, the branch, then passes through a global average Chi Huahou, activates by a sigmoid
Output and label 2 are sent into Euclidean distance Loss by layer output；Another branch is used to cluster feature, also passes through
One convolution module 2 is exported using a softmax active coating later by a full articulamentum, the output and label 1
It is sent into cross entropy Loss；
F, face quality evaluation: during face quality evaluation, we obtain two values that deep neural network provides,
First is that the mass fraction of network judgement, another is quality category probability；Final face quality is by the combination of the two come really
Fixed, score=a*b* classification correspondence score value+(1-a*b) * face mass fraction of face, wherein a is coefficient, and value range exists
[0,1], b be the face picture the corresponding probability of quality category maximum probability classification, classification correspond to score value be 10 classes correspondence [0,
1]。
Preferably, the data preprocessing method used in the step B mainly has brightness change, noise, Gaussian Blur, fortune
Dynamic model paste tone, is deviated, is blocked, rotating.
Preferably, the calculation of similarity uses COS distance in the D step.
The present invention have it is following the utility model has the advantages that
The quality evaluation of face of the present invention is applied to the screening before face alignment, is clustered with face characteristic, is more accorded with
Close the data shape that face alignment needs；Meanwhile this method is not necessarily to manually be labeled, using face characteristic sequence of calculation face
The similarity of feature is converted into trained label, carries out neural metwork training, process is simple, and speed is fast, can satisfy real-time and wants
It asks.
Detailed description of the invention
Fig. 1 is recognition of face flow chart in the prior art；
Fig. 2 is face Evaluation Model on Quality of the present invention training flow chart；
Fig. 3 is inventive network structure chart；
Fig. 4 is the flow chart of convolution module 1 of the present invention；
Fig. 5 is the flow chart of convolution module 2 of the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
Fig. 1-5 is please referred to, the present invention provides a kind of technical solution: a kind of face quality evaluation side based on feature clustering
Method, including the following steps:
A, data preparation: training data of the present invention uses CMU Multi-PIE face database, it is possible to use other
Different angle comprising different faces id, illumination, the human face photo of clarity；Multi-PIE data comprise more than 337id's
750000+ face Various Seasonal, different angle, the human face photo of different illumination；
B, it data prediction: before data are sent into training, needs to pre-process data, pretreated purpose one
It is in order to enrich sample number amount and type, second is that in order to avoid over-fitting；
C, feature extraction: face characteristic extraction algorithm is using the face recognition to increase income, it is possible to use other people
Face knows method for distinguishing, and the certificate photo (everyone includes 3-5 of Various Seasonal) of 337id people in data set Multi-PIE uses
Face recognition algorithm extracts face characteristic and forms gallery, and gallery data are extracted feature and located in advance without image
Module is managed, image preprocessing is passed through for other data, by the image zooming-out face characteristic after pretreatment；
D, training label generates: it will currently be extracted face characteristic and the id certificate photo feature calculation similarity, it is similar
The distance between two feature of degree=1-, obtains the similarity between two faces, which is divided into 10 classes, as the face figure
The label 1 of picture, similarity value are the label 2 of the picture；
E, face quality model training: the data after pretreatment and two labels generated are fed together shown in Fig. 3
Depth network, in Fig. 3 after the last one convolution module 1, network is divided into Liang Ge branch, and a branch is used to return face
A convolution module 2 is passed through by score value, the branch, then passes through a global average Chi Huahou, activates by a sigmoid
Output and label 2 are sent into Euclidean distance Loss by layer output；Another branch is used to cluster feature, also passes through
One convolution module 2 is exported using a softmax active coating later by a full articulamentum, the output and label 1
It is sent into cross entropy Loss；
F, face quality evaluation: during face quality evaluation, we obtain two values that deep neural network provides,
First is that the mass fraction of network judgement, another is quality category probability；Final face quality is by the combination of the two come really
Fixed, score=a*b* classification correspondence score value+(1-a*b) * face mass fraction of face, wherein a is coefficient, and value range exists
[0,1], b be the face picture the corresponding probability of quality category maximum probability classification, classification correspond to score value be 10 classes correspondence [0,
1]。
Further, the data preprocessing method used in the step B mainly have brightness change, noise, Gaussian Blur,
Motion blur tone, is deviated, is blocked, rotating.
Further, the calculation of similarity uses COS distance in the D step.
It should be noted that, in this document, relational terms such as first and second and the like are used merely to a reality
Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation
In any actual relationship or order or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to
Non-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including those
Element, but also including other elements that are not explicitly listed, or further include for this process, method, article or equipment
Intrinsic element.
It although an embodiment of the present invention has been shown and described, for the ordinary skill in the art, can be with
A variety of variations, modification, replacement can be carried out to these embodiments without departing from the principles and spirit of the present invention by understanding
And modification, the scope of the present invention is defined by the appended.
Claims (3)
1. a kind of face method for evaluating quality based on feature clustering, which is characterized in that including the following steps:
A, data preparation: training data of the present invention uses CMU Multi-PIE face database, it is possible to use other include
The different angle of different faces id, illumination, the human face photo of clarity；Multi-PIE data comprise more than the 750000 of 337id
+ face Various Seasonal, different angle, the human face photo of different illumination；
B, it data prediction: before data are sent into training, needs to pre-process data, pretreated purpose is first is that be
Abundant sample number amount and type, second is that in order to avoid over-fitting；
C, feature extraction: face characteristic extraction algorithm is using the face recognition to increase income, it is possible to use other faces are known
The certificate photo (everyone includes 3-5 of Various Seasonal) of 337id people in data set Multi-PIE is used face by method for distinguishing
Recognition algorithm extracts face characteristic and forms gallery, and gallery data extract feature without image preprocessing mould
Block passes through image preprocessing for other data, by the image zooming-out face characteristic after pretreatment；
D, training label generates: it will currently be extracted face characteristic and the id certificate photo feature calculation similarity, similarity=
The distance between two feature of 1-, obtains the similarity between two faces, which is divided into 10 classes, as the facial image
Label 1, similarity value are the label 2 of the picture；
E, the data after pretreatment and two labels generated face quality model training: are fed together depth shown in Fig. 3
Network is spent, in Fig. 3 after the last one convolution module 1, network is divided into Liang Ge branch, and a branch is used to return point of face
A convolution module 2 is passed through by value, the branch, then passes through a global average Chi Huahou, by a sigmoid active coating
Output and label 2 are sent into Euclidean distance Loss by output；Another branch is used to cluster feature, also passes through one
A convolution module 2 is exported using a softmax active coating later by a full articulamentum, which send with label 1
Enter cross entropy Loss；
F, face quality evaluation: during face quality evaluation, we obtain two values that deep neural network provides, first is that
The mass fraction of network judgement, another is quality category probability；Finally face quality is determined by the combination of the two, people
The score of face=a*b* classification correspondence score value+(1-a*b) * face mass fraction, wherein a be coefficient, value range [0,
1], b is the corresponding probability of quality category maximum probability classification of the face picture, and it is that 10 classes are corresponding [0,1] that classification, which corresponds to score value,.
2. a kind of face method for evaluating quality based on feature clustering according to claim 1, it is characterised in that: the B
The data preprocessing method used in step mainly has brightness change, noise, Gaussian Blur, motion blur, tone, offset, screening
Gear, rotation etc..
3. a kind of face method for evaluating quality based on feature clustering according to claim 1, it is characterised in that: the D
The calculation of similarity uses COS distance in step.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910715538.3A CN110427888A (en) | 2019-08-05 | 2019-08-05 | A kind of face method for evaluating quality based on feature clustering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910715538.3A CN110427888A (en) | 2019-08-05 | 2019-08-05 | A kind of face method for evaluating quality based on feature clustering |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110427888A true CN110427888A (en) | 2019-11-08 |
Family
ID=68412520
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910715538.3A Pending CN110427888A (en) | 2019-08-05 | 2019-08-05 | A kind of face method for evaluating quality based on feature clustering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110427888A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111401344A (en) * | 2020-06-04 | 2020-07-10 | 腾讯科技（深圳）有限公司 | Face recognition method and device and training method and device of face recognition system |
| CN111696090A (en) * | 2020-06-08 | 2020-09-22 | 电子科技大学 | Method for evaluating quality of face image in unconstrained environment |
| CN112215822A (en) * | 2020-10-13 | 2021-01-12 | 北京中电兴发科技有限公司 | Face image quality evaluation method based on lightweight regression network |
| CN112948612A (en) * | 2021-03-16 | 2021-06-11 | 杭州海康威视数字技术股份有限公司 | Human body cover generation method and device, electronic equipment and storage medium |
| CN114155589A (en) * | 2021-11-30 | 2022-03-08 | 北京百度网讯科技有限公司 | Image processing method, device, equipment and storage medium |
| CN115512427A (en) * | 2022-11-04 | 2022-12-23 | 北京城建设计发展集团股份有限公司 | User face registration method and system combined with matched biopsy |
| CN118197609A (en) * | 2024-05-17 | 2024-06-14 | 大连百首企家科技有限公司 | Anesthesia and analgesia effect evaluation method based on facial expression analysis |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140044348A1 (en) * | 2011-02-18 | 2014-02-13 | National Ict Australia Limited | Image quality assessment |
| CN104408736A (en) * | 2014-12-12 | 2015-03-11 | 西安电子科技大学 | Characteristic-similarity-based synthetic face image quality evaluation method |
| US20160086015A1 (en) * | 2007-01-09 | 2016-03-24 | Si Corporation | Method and system for automated face detection and recognition |
| CN106897748A (en) * | 2017-03-02 | 2017-06-27 | 上海极链网络科技有限公司 | Face method for evaluating quality and system based on deep layer convolutional neural networks |
| CN107133601A (en) * | 2017-05-13 | 2017-09-05 | 五邑大学 | A kind of pedestrian's recognition methods again that network image super-resolution technique is resisted based on production |
| CN107273510A (en) * | 2017-06-20 | 2017-10-20 | 广东欧珀移动通信有限公司 | Photo recommends method and Related product |
| CN107292813A (en) * | 2017-05-17 | 2017-10-24 | 浙江大学 | A kind of multi-pose Face generation method based on generation confrontation network |
| CN107609493A (en) * | 2017-08-25 | 2018-01-19 | 广州视源电子科技股份有限公司 | Method and device for optimizing human face image quality evaluation model |
| CN107704806A (en) * | 2017-09-01 | 2018-02-16 | 深圳市唯特视科技有限公司 | A kind of method that quality of human face image prediction is carried out based on depth convolutional neural networks |
| CN107832802A (en) * | 2017-11-23 | 2018-03-23 | 北京智芯原动科技有限公司 | Quality of human face image evaluation method and device based on face alignment |
| CN108765407A (en) * | 2018-05-31 | 2018-11-06 | 上海依图网络科技有限公司 | A kind of portrait picture quality determination method and device |
| CN108960087A (en) * | 2018-06-20 | 2018-12-07 | 中国科学院重庆绿色智能技术研究院 | A kind of quality of human face image appraisal procedure and system based on various dimensions evaluation criteria |
| CN109117797A (en) * | 2018-08-17 | 2019-01-01 | 浙江捷尚视觉科技股份有限公司 | A kind of face snapshot recognition method based on face quality evaluation |
| CN109544523A (en) * | 2018-11-14 | 2019-03-29 | 北京智芯原动科技有限公司 | Quality of human face image evaluation method and device based on more attribute face alignments |
-
2019
- 2019-08-05 CN CN201910715538.3A patent/CN110427888A/en active Pending
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160086015A1 (en) * | 2007-01-09 | 2016-03-24 | Si Corporation | Method and system for automated face detection and recognition |
| US20140044348A1 (en) * | 2011-02-18 | 2014-02-13 | National Ict Australia Limited | Image quality assessment |
| CN104408736A (en) * | 2014-12-12 | 2015-03-11 | 西安电子科技大学 | Characteristic-similarity-based synthetic face image quality evaluation method |
| CN106897748A (en) * | 2017-03-02 | 2017-06-27 | 上海极链网络科技有限公司 | Face method for evaluating quality and system based on deep layer convolutional neural networks |
| CN107133601A (en) * | 2017-05-13 | 2017-09-05 | 五邑大学 | A kind of pedestrian's recognition methods again that network image super-resolution technique is resisted based on production |
| CN107292813A (en) * | 2017-05-17 | 2017-10-24 | 浙江大学 | A kind of multi-pose Face generation method based on generation confrontation network |
| CN107273510A (en) * | 2017-06-20 | 2017-10-20 | 广东欧珀移动通信有限公司 | Photo recommends method and Related product |
| CN107609493A (en) * | 2017-08-25 | 2018-01-19 | 广州视源电子科技股份有限公司 | Method and device for optimizing human face image quality evaluation model |
| CN107704806A (en) * | 2017-09-01 | 2018-02-16 | 深圳市唯特视科技有限公司 | A kind of method that quality of human face image prediction is carried out based on depth convolutional neural networks |
| CN107832802A (en) * | 2017-11-23 | 2018-03-23 | 北京智芯原动科技有限公司 | Quality of human face image evaluation method and device based on face alignment |
| CN108765407A (en) * | 2018-05-31 | 2018-11-06 | 上海依图网络科技有限公司 | A kind of portrait picture quality determination method and device |
| CN108960087A (en) * | 2018-06-20 | 2018-12-07 | 中国科学院重庆绿色智能技术研究院 | A kind of quality of human face image appraisal procedure and system based on various dimensions evaluation criteria |
| CN109117797A (en) * | 2018-08-17 | 2019-01-01 | 浙江捷尚视觉科技股份有限公司 | A kind of face snapshot recognition method based on face quality evaluation |
| CN109544523A (en) * | 2018-11-14 | 2019-03-29 | 北京智芯原动科技有限公司 | Quality of human face image evaluation method and device based on more attribute face alignments |
Non-Patent Citations (7)
| Title |
|---|
| JAVIER HERNANDEZ-ORTEGA 等: "FaceQnet: Quality Assessment for Face Recognition based on Deep Learning", 《ARXIV》 * |
| LACEY BEST-ROWDEN 等: "Automatic Face Image Quality Prediction", 《ARXIV》 * |
| RALPH GROSS 等: "Multi-PIE", 《PREPRINT SUBMITTED TO IMAGE AND VISION COMPUTING》 * |
| VISHAL AGARWAL: "Deep Face Quality Assessment", 《ARXIV》 * |
| YEZHOU LI 等: "Image quality assessment using deep convolutional networks", 《AIP ADVANCES》 * |
| 李秋珍 等: "基于卷积神经网络的人脸图像质量评价", 《计算机应用》 * |
| 程换新 等: "基于迁移学习的人脸图像质量评估", 《电子测量技术》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111401344A (en) * | 2020-06-04 | 2020-07-10 | 腾讯科技（深圳）有限公司 | Face recognition method and device and training method and device of face recognition system |
| CN111401344B (en) * | 2020-06-04 | 2020-09-29 | 腾讯科技（深圳）有限公司 | Face recognition method and device and training method and device of face recognition system |
| CN111696090A (en) * | 2020-06-08 | 2020-09-22 | 电子科技大学 | Method for evaluating quality of face image in unconstrained environment |
| CN112215822A (en) * | 2020-10-13 | 2021-01-12 | 北京中电兴发科技有限公司 | Face image quality evaluation method based on lightweight regression network |
| CN112948612A (en) * | 2021-03-16 | 2021-06-11 | 杭州海康威视数字技术股份有限公司 | Human body cover generation method and device, electronic equipment and storage medium |
| CN112948612B (en) * | 2021-03-16 | 2024-02-06 | 杭州海康威视数字技术股份有限公司 | Human body cover generation method and device, electronic equipment and storage medium |
| CN114155589A (en) * | 2021-11-30 | 2022-03-08 | 北京百度网讯科技有限公司 | Image processing method, device, equipment and storage medium |
| CN114155589B (en) * | 2021-11-30 | 2023-08-08 | 北京百度网讯科技有限公司 | Image processing method, device, equipment and storage medium |
| CN115512427A (en) * | 2022-11-04 | 2022-12-23 | 北京城建设计发展集团股份有限公司 | User face registration method and system combined with matched biopsy |
| CN115512427B (en) * | 2022-11-04 | 2023-04-25 | 北京城建设计发展集团股份有限公司 | User face registration method and system combined with matched biopsy |
| CN118197609A (en) * | 2024-05-17 | 2024-06-14 | 大连百首企家科技有限公司 | Anesthesia and analgesia effect evaluation method based on facial expression analysis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110427888A (en) | A kind of face method for evaluating quality based on feature clustering | |
| CN108520216B (en) | Gait image-based identity recognition method | |
| CN106875381A (en) | A kind of phone housing defect inspection method based on deep learning | |
| CN108537136A (en) | The pedestrian's recognition methods again generated based on posture normalized image | |
| CN107506702A (en) | Human face recognition model training and test system and method based on multi-angle | |
| CN107967695A (en) | A kind of moving target detecting method based on depth light stream and morphological method | |
| CN103295009B (en) | Based on the license plate character recognition method of Stroke decomposition | |
| CN112766218B (en) | Cross-domain pedestrian re-recognition method and device based on asymmetric combined teaching network | |
| CN110472652A (en) | A small amount of sample classification method based on semanteme guidance | |
| CN108960201A (en) | A kind of expression recognition method extracted based on face key point and sparse expression is classified | |
| CN113298018A (en) | False face video detection method and device based on optical flow field and facial muscle movement | |
| CN113205107A (en) | Vehicle type recognition method based on improved high-efficiency network | |
| CN106295532A (en) | A kind of human motion recognition method in video image | |
| CN107220598A (en) | Iris Texture Classification based on deep learning feature and Fisher Vector encoding models | |
| CN110503078A (en) | A kind of remote face identification method and system based on deep learning | |
| CN111639580A (en) | Gait recognition method combining feature separation model and visual angle conversion model | |
| CN111738178A (en) | Wearing mask facial expression recognition method based on deep learning | |
| CN114842524A (en) | Face false distinguishing method based on irregular significant pixel cluster | |
| CN116030396A (en) | Accurate segmentation method for video structured extraction | |
| CN109165542A (en) | Pedestrian detection method based on simplified convolutional neural network | |
| Tong et al. | Research on face recognition method based on deep neural network | |
| CN112488165A (en) | Infrared pedestrian identification method and system based on deep learning model | |
| CN111950452A (en) | Face recognition method | |
| CN117036412A (en) | Twin network infrared pedestrian target tracking method integrating deformable convolution | |
| CN109145744B (en) | LSTM network pedestrian re-identification method based on self-adaptive prediction mode |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191108 |
|
| RJ01 | Rejection of invention patent application after publication |