CN106529377A - Age estimating method, age estimating device and age estimating system based on image - Google Patents
Age estimating method, age estimating device and age estimating system based on image Download PDFInfo
- Publication number
- CN106529377A CN106529377A CN201510586848.1A CN201510586848A CN106529377A CN 106529377 A CN106529377 A CN 106529377A CN 201510586848 A CN201510586848 A CN 201510586848A CN 106529377 A CN106529377 A CN 106529377A
- Authority
- CN
- China
- Prior art keywords
- sample
- face
- detected
- image
- age
- 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 OR CALCULATING; 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
- G06V40/171—Local features and components; Facial parts ; Occluding parts, e.g. glasses; Geometrical relationships
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
- G06V40/178—Human faces, e.g. facial parts, sketches or expressions estimating age from face image; using age information for improving recognition
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)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses an age estimating method, an age estimating device and an age estimating system based on an image, wherein the age estimating method based on the image comprises the steps of acquiring a to-be-detected face image and corresponding key points, wherein the key points are angular point positions of main parts; extracting an original characteristic of the to-be-detected face according to the to-be-detected face image and corresponding key points; performing dimension reduction processing on the original characteristic of the to-be-detected face, and acquiring a low-dimension characteristic vector of the to-be-detected face; and performing age estimation on the low-dimension characteristic vector of the to-be-detected face through an age estimating model. According to the age estimating method, the age estimating device and the age estimating system, through a face representation method for performing dimension reduction processing on the original characteristic, and the age estimating model, an average estimation error which is lower than 5.5 can be realized in dynamic scene application in age estimation, and an accuracy in a range in which an absolute error is lower than ten years reaches 88%. Furthermore, training of the age estimating model does not require identity information of samples, thereby greatly facilitating acquisition of training samples.
Description
Technical Field
The invention relates to the technical field of computer biological identification, in particular to an age estimation method, device and system based on images.
Background
With the intensive research of face recognition technology, age estimation based on face images has become one of the popular research subjects in the field of computer biometrics. The way age estimation or verification is currently undertaken in most cases depends on subjective evaluation of the person or on the relevant document, such as a passport, an identification card, etc. These methods have the disadvantages of slow speed, high cost, unfriendliness, easy counterfeiting and the like. Based on the defects, the prior art adopts face recognition technology to realize age estimation, so that many places where the original age estimation is not easy to implement can be covered by the technology, such as the places where the age distribution of store personnel is analyzed in market monitoring, automatic service terminals for providing special services for users of different ages and the like, and the technology has wide market prospect.
The method for realizing age estimation through the face recognition technology in the prior art is realized by the following specific steps: firstly, intercepting a face area to be detected; secondly, detecting key points of the face area; thirdly, extracting a characteristic vector of the face image to be estimated to generate a candidate age growing mode vector; fourthly, judging whether the growth mode subspace is trained well; training a growth mode subspace if the growth mode subspace is not trained; if the growth mode subspace is trained, entering the next step; fifthly, projecting all candidate age growing mode vectors of the face image to be estimated into a growing mode subspace according to the trained growing mode subspace, and reconstructing complete candidate age growing mode vectors from the projection vectors; and sixthly, finding out an optimal age growing mode by comparing the reconstruction error between the reconstructed image and the original image, and estimating the current age of the human face image to be estimated at the position of the optimal age growing mode.
Therefore, in the process of designing a method for realizing age estimation by using a face recognition technology, the inventor finds that at least the following problems exist in the prior art:
the above-mentioned face recognition technology realizes the disadvantage of the age estimation scheme: first, the face representation is a less than complete representation that does not adequately represent age-related information in the face image; secondly, training of the growth mode subspace requires that the sample simultaneously marks two kinds of information, namely age and identity, and under the normal condition, the sample is difficult to obtain in large quantity; third, age estimation requires projection and reconstruction of nearly a hundred candidate growth mode vectors to find the growth mode with the least reconstruction error, a process that is tedious and time consuming.
Disclosure of Invention
In view of the above problems, the present invention is proposed to provide a solution to overcome or at least partially solve the above problems, and the technical solution of the present invention is realized by:
in one aspect, the present invention provides an age estimation method based on an image, including:
acquiring a face image to be detected and corresponding key points; the key points are the angular point positions of the main parts;
extracting original features of the face to be detected according to the face image to be detected and the corresponding key points;
carrying out dimensionality reduction on the original features of the face to be detected to obtain a low-dimensionality feature vector of the face to be detected;
and carrying out age estimation on the low-dimensional characteristic vector of the face to be detected through an age estimation model.
Preferably, the method further comprises:
the step of extracting the original features of the face to be detected according to the face image to be detected and the corresponding key points specifically comprises the following steps:
acquiring local features of the face to be detected according to the face image to be detected and the corresponding key points;
the local features of the face to be detected are connected in series to obtain original features of the face to be detected;
the step of performing dimensionality reduction on the original features of the face to be detected to obtain low-dimensionality feature vectors of the face to be detected specifically comprises the following steps:
acquiring a feature dimension reduction matrix and the original features of the face to be detected;
and performing dimension reduction processing on the original features of the face to be detected through the feature dimension reduction matrix to obtain a low-dimensional feature vector of the face to be detected.
Preferably, the method further comprises:
acquiring a face image sample information set and the age of a corresponding image sample;
determining a key point of each sample in the face image sample information set; the key points are the angular point positions of the main parts;
acquiring the original characteristics of each sample in the face image sample information set according to each sample in the face image sample information set and key points of each sample;
carrying out dimensionality reduction on the original features of each sample in the face image sample information set to obtain a low-dimensional feature vector of each sample;
and training an age estimation model through the low-dimensional characteristic vector of each sample and the corresponding age value of the low-dimensional characteristic vector, and acquiring the age estimation model for subsequent age estimation.
Preferably, the step of obtaining the original features of each sample in the face image sample information set according to each sample in the face image sample information set and the key point thereof includes:
obtaining the local characteristics of each sample according to the key points of each sample in the face image sample information set;
and connecting the local features of the same sample image in the face image sample information set in series to obtain the original feature of each sample.
Preferably, the step of performing dimension reduction processing on the original feature of each sample in the face image sample information set to obtain a low-dimensional feature vector of each sample includes:
obtaining a characteristic dimension reduction matrix through a dimension reduction algorithm;
and performing dimensionality reduction treatment on the original features of each sample through the feature dimensionality reduction matrix to obtain a low-dimensionality feature vector of each sample.
Preferably, the training age estimation model is constrained by a penalty function;
the penalty function is:
wherein,x(i)for the output of the ith sample at the last hidden layer of the multi-layer neural network of the output layer, y(i)Is the actual age of the ith sample, m is the number of training samples, k is the number of age classes, wijRepresents the confidence that the ith sample is age j, when(i)When the-j | < K,in other cases, wij＝0。
In another aspect, the present invention provides an image-based age estimation apparatus, including:
the information acquisition unit is used for acquiring a face image to be detected and corresponding key points; the key points are the angular point positions of the main parts;
the feature extraction unit is used for extracting the original features of the face to be detected according to the face image to be detected and the corresponding key points;
the dimension reduction unit is used for carrying out dimension reduction processing on the original features of the face to be detected to obtain a low-dimensional feature vector of the face to be detected;
and the age estimation unit is used for carrying out age estimation on the low-dimensional characteristic vector of the face to be detected through an age estimation model.
Preferably, the feature extraction unit specifically includes:
the local feature acquisition subunit is used for acquiring local features of the face to be detected according to the face image to be detected and the corresponding key points;
the original characteristic obtaining subunit is used for connecting the local characteristics of the face to be detected in series to obtain original characteristics of the face to be detected;
the dimension reduction unit is also used for acquiring a feature dimension reduction matrix and the original features of the face to be detected; and performing dimension reduction processing on the original features of the face to be detected through the feature dimension reduction matrix to obtain a low-dimensional feature vector of the face to be detected.
Preferably, the apparatus further comprises:
the system comprises a sample information acquisition unit, a face image acquisition unit and a face image acquisition unit, wherein the sample information acquisition unit is used for acquiring a face image sample information set and the age of a corresponding image sample;
the position determining unit is used for determining the key point of each sample in the face image sample information set; the key points are the angular point positions of the main parts;
the original sample characteristic acquisition unit is used for acquiring the original characteristic of each sample in the face image sample information set according to each sample in the face image sample information set and the key point of each sample;
the sample dimension reduction unit is used for carrying out dimension reduction processing on the original features of each sample in the face image sample information set to obtain a low-dimensional feature vector of each sample;
and the model acquisition unit is used for training an age estimation model through the low-dimensional characteristic vector of each sample and the corresponding age value of each sample, and acquiring the age estimation model for subsequent age estimation.
Preferably, the sample original feature obtaining unit is configured to obtain a local feature of each sample according to a key point of each sample in the face image sample information set; and connecting the local features of the same sample image in the face image sample information set in series to obtain the original feature of each sample.
Preferably, the sample dimension reduction unit is configured to obtain a feature dimension reduction matrix through a dimension reduction algorithm; and performing dimensionality reduction treatment on the original features of each sample through the feature dimensionality reduction matrix to obtain a low-dimensionality feature vector of each sample.
In yet another aspect, the present invention provides an image-based age estimation system, comprising: the image-based age estimation apparatus as claimed in any one of the above.
According to the invention, through the face representation method for processing the original characteristics by dimension reduction and the age estimation model, the age estimation can reach an average estimation error smaller than 5.5 in the application of a dynamic scene, and the absolute error is smaller than the accuracy of 10 years old.
Drawings
Fig. 1 is a flowchart of an age estimation method based on images according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of an age estimation apparatus based on images according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of an age estimation system based on images according to an embodiment of the present invention;
fig. 4 is a flowchart illustrating an age estimation model training process in an image-based age estimation method according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in detail with reference to the accompanying drawings.
Fig. 1 is a schematic diagram illustrating an age estimation method based on images according to an embodiment of the present invention; the method comprises the following steps:
101: acquiring a face image to be detected and corresponding key points; the key points are the angular point positions of the main parts;
102: extracting original features of the face to be detected according to the face image to be detected and the corresponding key points;
103: carrying out dimensionality reduction on the original features of the face to be detected to obtain a low-dimensionality feature vector of the face to be detected;
104: and carrying out age estimation on the low-dimensional characteristic vector of the face to be detected through an age estimation model.
It should be noted that the step of extracting the original features of the face to be detected according to the face image to be detected and the corresponding key points specifically includes:
acquiring local features of the face to be detected according to the face image to be detected and the corresponding key points;
the local features of the face to be detected are connected in series to obtain original features of the face to be detected;
the step of performing dimensionality reduction on the original features of the face to be detected to obtain low-dimensionality feature vectors of the face to be detected specifically comprises the following steps:
acquiring a feature dimension reduction matrix and the original features of the face to be detected;
and performing dimension reduction processing on the original features of the face to be detected through the feature dimension reduction matrix to obtain a low-dimensional feature vector of the face to be detected.
Based on the above embodiments, fig. 4 is a flowchart illustrating the training of an age estimation model in an image-based age estimation method; the training process is as follows:
401: acquiring a face image sample information set and the age of a corresponding image sample;
402: determining a key point of each sample in the face image sample information set; the key points are the angular point positions of the main parts; it should be noted that the key points may also be directly obtained;
403: acquiring the original characteristics of each sample in the face image sample information set according to each sample in the face image sample information set and key points of each sample;
404: carrying out dimensionality reduction on the original features of each sample in the face image sample information set to obtain a low-dimensional feature vector of each sample;
405: and training an age estimation model through the low-dimensional characteristic vector of each sample and the corresponding age value of the low-dimensional characteristic vector, and acquiring the age estimation model for subsequent age estimation.
Preferably, the step of obtaining the original features of each sample in the face image sample information set according to each sample in the face image sample information set and the key point thereof includes:
obtaining the local characteristics of each sample according to the key points of each sample in the face image sample information set;
and connecting the local features of the same sample image in the face image sample information set in series to obtain the original feature of each sample.
Preferably, the step of performing dimension reduction processing on the original feature of each sample in the face image sample information set to obtain a low-dimensional feature vector of each sample includes:
obtaining a characteristic dimension reduction matrix through a dimension reduction algorithm;
and performing dimensionality reduction treatment on the original features of each sample through the feature dimensionality reduction matrix to obtain a low-dimensionality feature vector of each sample.
Preferably, the training age estimation model is constrained by a penalty function;
the penalty function is:
wherein,x(i)for the output of the ith sample at the last hidden layer of the multi-layer neural network of the output layer, y(i)Is the actual age of the ith sample, m is the number of training samples, k is the number of age classes, wijRepresents the confidence that the ith sample is age j, when(i)When the-j | < K,in other cases, wij＝0。
Based on the above embodiment, as shown in fig. 2, a schematic structural diagram of an age estimation device based on an image according to an embodiment of the present invention is provided; the device includes:
an information obtaining unit 201, configured to obtain a face image to be detected and corresponding key points; the key points are the angular point positions of the main parts;
a feature extraction unit 202, configured to extract an original feature of the face to be detected according to the face image to be detected and the corresponding key point;
the dimension reduction processing unit 203 is configured to perform dimension reduction processing on the original features of the face to be detected to obtain a low-dimensional feature vector of the face to be detected;
and the age estimation unit 204 is configured to perform age estimation on the low-dimensional feature vector of the face to be detected through an age estimation model.
Preferably, the feature extraction unit specifically includes:
the local feature acquisition subunit is used for acquiring local features of the face to be detected according to the face image to be detected and the corresponding key points;
the original characteristic obtaining subunit is used for connecting the local characteristics of the face to be detected in series to obtain original characteristics of the face to be detected;
the dimension reduction unit is also used for acquiring a feature dimension reduction matrix and the original features of the face to be detected; and performing dimension reduction processing on the original features of the face to be detected through the feature dimension reduction matrix to obtain a low-dimensional feature vector of the face to be detected.
Preferably, the apparatus further comprises:
the system comprises a sample information acquisition unit, a face image acquisition unit and a face image acquisition unit, wherein the sample information acquisition unit is used for acquiring a face image sample information set and the age of a corresponding image sample;
the position determining unit is used for determining the key point of each sample in the face image sample information set; the key points are the angular point positions of the main parts;
the original sample characteristic acquisition unit is used for acquiring the original characteristic of each sample in the face image sample information set according to each sample in the face image sample information set and the key point of each sample;
the sample dimension reduction unit is used for carrying out dimension reduction processing on the original features of each sample in the face image sample information set to obtain a low-dimensional feature vector of each sample;
and the model acquisition unit is used for training an age estimation model through the low-dimensional characteristic vector of each sample and the corresponding age value of each sample, and acquiring the age estimation model for subsequent age estimation.
Preferably, the sample original feature obtaining unit is configured to obtain a local feature of each sample according to a key point of each sample in the face image sample information set; and connecting the local features of the same sample image in the face image sample information set in series to obtain the original feature of each sample.
Preferably, the sample dimension reduction unit is configured to obtain a feature dimension reduction matrix through a dimension reduction algorithm; and performing dimensionality reduction treatment on the original features of each sample through the feature dimensionality reduction matrix to obtain a low-dimensionality feature vector of each sample.
Based on the above embodiments, the training principle of the age estimation model and the age estimation principle are described in detail below.
The training principle of the age estimation model is specifically realized as follows:
firstly, acquiring a face image sample information set and the age of a corresponding image sample;
secondly, determining key points of each sample in the face image sample information set; the key points can also be directly acquired through the first step at the same time;
thirdly, local features are calculated according to each sample and key points thereof in the face image sample information set; and connecting the local features of the same sample image in series to obtain the original features. The original features are a complete representation of a human face.
Because the human face is not normalized on the whole, the problems caused by factors such as posture, expression and facial form difference can be effectively avoided, and meanwhile, the local features extracted on the local area of the key points can also obtain more precise and complete human face representation, which is beneficial to the final age estimation. The original feature description can adopt HOG, LBP, Gabor and the like, and the size, gridding parameters and the like of each key point local area can be freely set according to actual conditions.
Fourthly, performing dimensionality reduction on the original features; in the feature dimension reduction processing, PCA, LDA and the like can be used as the dimension reduction method. The original human face features obtained in the third step are usually high in dimensionality, feature dimensionality can be reduced through dimensionality reduction, and a noise reduction effect is achieved, so that subsequent processing is facilitated. The dimension reduction process is described below by taking the PCA dimension reduction method as an example: the method comprises the following two steps: 1, obtaining a characteristic dimension reduction matrix through a PCA algorithm, wherein the characteristic dimension reduction matrix is a dimension reduction model which is only obtained in a training process of an age estimation model; 2. and performing dimension reduction processing on the original features through the feature dimension reduction matrix to obtain low-dimension feature vectors.
And fifthly, training an age estimation model through the low-dimensional characteristic vector of each sample and the corresponding age value of the sample to obtain the age estimation model for subsequent age estimation.
It should be noted that, a penalty function is adopted for constraint in the process of the age estimation model; the penalty function is as follows:
wherein,x(i)for the output of the ith sample at the last hidden layer of the multi-layer neural network of the output layer, y(i)Is the actual age of the ith sample, m is the number of training samples, k is the number of age classes, wijRepresents the confidence that the ith sample is age j, when(i)When the-j | < K,in other cases, wij＝0。
The age estimation process of the image-based age estimation method is specifically as follows:
firstly, acquiring a face image to be detected and corresponding key points; the key points are the angular point positions of the main parts; for example: the angular point positions of the eyes, nose, mouth, etc.
Secondly, extracting the original features of the face to be detected according to the detected face image and the corresponding key points; the method specifically comprises the following steps:
acquiring local features of the face to be detected according to the face image to be detected and the corresponding key points;
the local features of the face to be detected are connected in series to obtain original features of the face to be detected;
thirdly, performing dimensionality reduction on the original features of the face to be detected to obtain a low-dimensionality feature vector of the face to be detected; the method specifically comprises the following steps:
acquiring a feature dimension reduction matrix and the original features of the face to be detected;
and performing dimension reduction processing on the original features of the face to be detected through the feature dimension reduction matrix to obtain a low-dimensional feature vector of the face to be detected.
And fourthly, carrying out age estimation on the low-dimensional characteristic vector of the face to be detected through an age estimation model.
It should be noted that the age estimation model is a multilayer neural network; the multilayer neural network outputs the attribute information of the face to be detected according to the probability that the current sample belongs to each age value; and taking the age value with the maximum probability output as the final age estimation value of the face to be detected.
It should be noted that the present invention aims to estimate the age based on a face image, and the age estimation is a complex concept relating to physiological, sociological, and psychological problems, and is not equivalent to the physiological age, the category thereof is fuzzy, and the risk of generating an erroneous estimation is unbalanced, and the feeling of misestimating a child of 10 years to 15 years and 50 years brings a great difference to the user, so that the age estimation cannot be simply regarded as a multi-classification problem. Aiming at the situations, the invention designs a multilayer neural network taking Softmax Regression as an output layer, directly judges the probability that the sample belongs to different age values, and designs a special penalty function aiming at the imbalance of the error estimation risk in the training process of the age estimation model. The penalty function is
x(i)For the output of the ith sample at the last hidden layer in the network, y(i)Is the actual age of the ith sample, m is the number of training samples, k is the number of age classes, wijRepresents the confidence that the ith sample is age j, when(i)When the-j | < K,in other cases, wij＝0。
The network training adopts a back propagation gradient descent algorithm.
Fig. 3 is a schematic structural diagram of an age estimation system based on images according to an embodiment of the present invention; the system comprises: an image-based age estimation apparatus as described in any one of the above.
According to the invention, an accurate age estimation system is established through original features of dimension reduction processing, a fine and complete face representation method and a specially designed classifier aiming at the age estimation problem, the system can achieve an average estimation error smaller than 5.5 in the application of a dynamic scene, and the accuracy rate of an absolute error smaller than 10 years old reaches 88%. In addition, the training of the classifier does not need to know the identity information of the sample, thereby greatly facilitating the acquisition of the training sample.
The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.
Claims (12)
1. An image-based age estimation method, comprising:
acquiring a face image to be detected and corresponding key points; the key points are the angular point positions of the main parts;
extracting original features of the face to be detected according to the face image to be detected and the corresponding key points;
carrying out dimensionality reduction on the original features of the face to be detected to obtain a low-dimensionality feature vector of the face to be detected;
and carrying out age estimation on the low-dimensional characteristic vector of the face to be detected through an age estimation model.
2. The image-based age estimation method according to claim 1,
the step of extracting the original features of the face to be detected according to the face image to be detected and the corresponding key points specifically comprises the following steps:
acquiring local features of the face to be detected according to the face image to be detected and the corresponding key points;
the local features of the face to be detected are connected in series to obtain original features of the face to be detected;
the step of performing dimensionality reduction on the original features of the face to be detected to obtain low-dimensionality feature vectors of the face to be detected specifically comprises the following steps:
acquiring a feature dimension reduction matrix and the original features of the face to be detected;
and performing dimension reduction processing on the original features of the face to be detected through the feature dimension reduction matrix to obtain a low-dimensional feature vector of the face to be detected.
3. The image-based age estimation method according to claim 1 or 2, further comprising:
acquiring a face image sample information set and the age of a corresponding image sample;
determining a key point of each sample in the face image sample information set; the key points are the angular point positions of the main parts;
acquiring the original characteristics of each sample in the face image sample information set according to each sample in the face image sample information set and key points of each sample;
carrying out dimensionality reduction on the original features of each sample in the face image sample information set to obtain a low-dimensional feature vector of each sample;
and training an age estimation model through the low-dimensional characteristic vector of each sample and the corresponding age value of the low-dimensional characteristic vector, and acquiring the age estimation model for subsequent age estimation.
4. The image-based age estimation method according to claim 3, wherein the step of obtaining the original features of each sample in the face image sample information set according to each sample in the face image sample information set and its key points comprises:
obtaining the local characteristics of each sample according to the key points of each sample in the face image sample information set;
and connecting the local features of the same sample image in the face image sample information set in series to obtain the original feature of each sample.
5. The image-based age estimation method according to claim 4, wherein the step of performing a dimensionality reduction process on the original features of each sample in the face image sample information set to obtain a low-dimensional feature vector of each sample comprises:
obtaining a characteristic dimension reduction matrix through a dimension reduction algorithm;
and performing dimensionality reduction treatment on the original features of each sample through the feature dimensionality reduction matrix to obtain a low-dimensionality feature vector of each sample.
6. The image-based age estimation method of claim 5, wherein the trained age estimation model is constrained with a penalty function;
the penalty function is:
wherein,x(i)for the output of the ith sample at the last hidden layer of the multi-layer neural network of the output layer, y(i)Is the actual age of the ith sample, m is the number of training samples, k is the number of age classes, wijRepresents the confidence that the ith sample is age j, when(i)When the-j | < K,in other cases, wij＝0。
7. An image-based age estimation device, comprising:
the information acquisition unit is used for acquiring a face image to be detected and corresponding key points; the key points are the angular point positions of the main parts;
the feature extraction unit is used for extracting the original features of the face to be detected according to the face image to be detected and the corresponding key points;
the dimension reduction unit is used for carrying out dimension reduction processing on the original features of the face to be detected to obtain a low-dimensional feature vector of the face to be detected;
and the age estimation unit is used for carrying out age estimation on the low-dimensional characteristic vector of the face to be detected through an age estimation model.
8. The image-based age estimation apparatus according to claim 7,
the feature extraction unit specifically includes:
the local feature acquisition subunit is used for acquiring local features of the face to be detected according to the face image to be detected and the corresponding key points;
the original characteristic obtaining subunit is used for connecting the local characteristics of the face to be detected in series to obtain original characteristics of the face to be detected;
the dimension reduction unit is also used for acquiring a feature dimension reduction matrix and the original features of the face to be detected; and performing dimension reduction processing on the original features of the face to be detected through the feature dimension reduction matrix to obtain a low-dimensional feature vector of the face to be detected.
9. The image-based age estimation apparatus according to claim 7 or 8, further comprising:
the system comprises a sample information acquisition unit, a face image acquisition unit and a face image acquisition unit, wherein the sample information acquisition unit is used for acquiring a face image sample information set and the age of a corresponding image sample;
the position determining unit is used for determining the key point of each sample in the face image sample information set; the key points are the angular point positions of the main parts;
the original sample characteristic acquisition unit is used for acquiring the original characteristic of each sample in the face image sample information set according to each sample in the face image sample information set and the key point of each sample;
the sample dimension reduction unit is used for carrying out dimension reduction processing on the original features of each sample in the face image sample information set to obtain a low-dimensional feature vector of each sample;
and the model acquisition unit is used for training an age estimation model through the low-dimensional characteristic vector of each sample and the corresponding age value of each sample, and acquiring the age estimation model for subsequent age estimation.
10. The image-based age estimation device according to claim 9, wherein the sample original feature obtaining unit is configured to obtain a local feature of each sample according to a key point of each sample in the face image sample information set; and connecting the local features of the same sample image in the face image sample information set in series to obtain the original feature of each sample.
11. The image-based age estimation device according to claim 10, wherein the sample dimension reduction unit is configured to obtain a feature dimension reduction matrix through a dimension reduction algorithm; and performing dimensionality reduction treatment on the original features of each sample through the feature dimensionality reduction matrix to obtain a low-dimensionality feature vector of each sample.
12. An image-based age estimation system, comprising: an image-based age estimation apparatus as claimed in any one of claims 7 to 11.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510586848.1A CN106529377A (en) | 2015-09-15 | 2015-09-15 | Age estimating method, age estimating device and age estimating system based on image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510586848.1A CN106529377A (en) | 2015-09-15 | 2015-09-15 | Age estimating method, age estimating device and age estimating system based on image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106529377A true CN106529377A (en) | 2017-03-22 |
Family
ID=58349228
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510586848.1A Pending CN106529377A (en) | 2015-09-15 | 2015-09-15 | Age estimating method, age estimating device and age estimating system based on image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106529377A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107180234A (en) * | 2017-06-01 | 2017-09-19 | 四川新网银行股份有限公司 | The credit risk forecast method extracted based on expression recognition and face characteristic |
| CN107403145A (en) * | 2017-07-14 | 2017-11-28 | 北京小米移动软件有限公司 | Image characteristic points positioning method and device |
| CN107545249A (en) * | 2017-08-30 | 2018-01-05 | 国信优易数据有限公司 | A kind of population ages' recognition methods and device |
| CN110378306A (en) * | 2019-07-25 | 2019-10-25 | 厦门美图之家科技有限公司 | Age prediction technique, device and image processing equipment |
| CN110674744A (en) * | 2019-09-24 | 2020-01-10 | 京东方科技集团股份有限公司 | Age recognition method, device and electronic device |
| CN111062995A (en) * | 2019-11-28 | 2020-04-24 | 重庆中星微人工智能芯片技术有限公司 | Method and device for generating face image, electronic equipment and computer readable medium |
| CN111967382A (en) * | 2020-08-14 | 2020-11-20 | 北京金山云网络技术有限公司 | Age estimation method, and training method and device of age estimation model |
| CN114219977A (en) * | 2021-11-10 | 2022-03-22 | 中国工商银行股份有限公司 | Age estimation method, system, electronic device and storage medium |
| TWI832660B (en) * | 2023-01-06 | 2024-02-11 | 宏碁股份有限公司 | Method and device for calculating normative distribution of visual acuity index |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101615248A (en) * | 2009-04-21 | 2009-12-30 | 华为技术有限公司 | Age estimation method, device and face recognition system |
| CN102637251A (en) * | 2012-03-20 | 2012-08-15 | 华中科技大学 | Face recognition method based on reference features |
| CN102663413A (en) * | 2012-03-09 | 2012-09-12 | 中盾信安科技(江苏)有限公司 | Multi-gesture and cross-age oriented face image authentication method |
| CN103544486A (en) * | 2013-10-31 | 2014-01-29 | 东南大学 | Human age estimation method based on self-adaptation sign distribution |
| CN103927518A (en) * | 2014-04-14 | 2014-07-16 | 中国华戎控股有限公司 | Facial feature extraction method for facial analysis system |
| CN104021384A (en) * | 2014-06-30 | 2014-09-03 | 深圳市创冠智能网络技术有限公司 | Face recognition method and device |
| CN104463190A (en) * | 2014-10-30 | 2015-03-25 | 华为技术有限公司 | Age estimation method and equipment |
| CN104504376A (en) * | 2014-12-22 | 2015-04-08 | 厦门美图之家科技有限公司 | Age classification method and system for face images |
-
2015
- 2015-09-15 CN CN201510586848.1A patent/CN106529377A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101615248A (en) * | 2009-04-21 | 2009-12-30 | 华为技术有限公司 | Age estimation method, device and face recognition system |
| CN102663413A (en) * | 2012-03-09 | 2012-09-12 | 中盾信安科技(江苏)有限公司 | Multi-gesture and cross-age oriented face image authentication method |
| CN102637251A (en) * | 2012-03-20 | 2012-08-15 | 华中科技大学 | Face recognition method based on reference features |
| CN103544486A (en) * | 2013-10-31 | 2014-01-29 | 东南大学 | Human age estimation method based on self-adaptation sign distribution |
| CN103927518A (en) * | 2014-04-14 | 2014-07-16 | 中国华戎控股有限公司 | Facial feature extraction method for facial analysis system |
| CN104021384A (en) * | 2014-06-30 | 2014-09-03 | 深圳市创冠智能网络技术有限公司 | Face recognition method and device |
| CN104463190A (en) * | 2014-10-30 | 2015-03-25 | 华为技术有限公司 | Age estimation method and equipment |
| CN104504376A (en) * | 2014-12-22 | 2015-04-08 | 厦门美图之家科技有限公司 | Age classification method and system for face images |
Non-Patent Citations (1)
| Title |
|---|
| 李红梅: "基于人脸图像的性别识别的研究与实现", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107180234A (en) * | 2017-06-01 | 2017-09-19 | 四川新网银行股份有限公司 | The credit risk forecast method extracted based on expression recognition and face characteristic |
| CN107403145A (en) * | 2017-07-14 | 2017-11-28 | 北京小米移动软件有限公司 | Image characteristic points positioning method and device |
| CN107403145B (en) * | 2017-07-14 | 2021-03-09 | 北京小米移动软件有限公司 | Image feature point positioning method and device |
| CN107545249A (en) * | 2017-08-30 | 2018-01-05 | 国信优易数据有限公司 | A kind of population ages' recognition methods and device |
| CN110378306A (en) * | 2019-07-25 | 2019-10-25 | 厦门美图之家科技有限公司 | Age prediction technique, device and image processing equipment |
| CN110674744A (en) * | 2019-09-24 | 2020-01-10 | 京东方科技集团股份有限公司 | Age recognition method, device and electronic device |
| CN111062995A (en) * | 2019-11-28 | 2020-04-24 | 重庆中星微人工智能芯片技术有限公司 | Method and device for generating face image, electronic equipment and computer readable medium |
| CN111062995B (en) * | 2019-11-28 | 2024-02-23 | 重庆中星微人工智能芯片技术有限公司 | Method, apparatus, electronic device and computer readable medium for generating face image |
| CN111967382A (en) * | 2020-08-14 | 2020-11-20 | 北京金山云网络技术有限公司 | Age estimation method, and training method and device of age estimation model |
| CN114219977A (en) * | 2021-11-10 | 2022-03-22 | 中国工商银行股份有限公司 | Age estimation method, system, electronic device and storage medium |
| TWI832660B (en) * | 2023-01-06 | 2024-02-11 | 宏碁股份有限公司 | Method and device for calculating normative distribution of visual acuity index |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106529377A (en) | Age estimating method, age estimating device and age estimating system based on image | |
| Angulu et al. | Age estimation via face images: a survey | |
| CN108520216B (en) | Gait image-based identity recognition method | |
| CN104751108B (en) | Face image recognition device and face image recognition method | |
| CN109543526B (en) | Real and fake facial paralysis recognition system based on depth difference feature | |
| KR102174595B1 (en) | System and method for identifying faces in unconstrained media | |
| CN103942577B (en) | Based on the personal identification method for establishing sample database and composite character certainly in video monitoring | |
| CN111931758B (en) | Face recognition method and device combining facial veins | |
| CN109101938B (en) | Multi-label age estimation method based on convolutional neural network | |
| CN106778496A (en) | Biopsy method and device | |
| CN104143079A (en) | Method and system for face attribute recognition | |
| CN109299690B (en) | A method that can improve the accuracy of video real-time face recognition | |
| Arigbabu et al. | Integration of multiple soft biometrics for human identification | |
| JP2005149506A (en) | Method and apparatus for automatic object recognition/collation | |
| KR20130048076A (en) | Face recognition apparatus and control method for the same | |
| CN104123543A (en) | Eyeball movement identification method based on face identification | |
| CN106529504B (en) | A kind of bimodal video feeling recognition methods of compound space-time characteristic | |
| CN104598888B (en) | A kind of recognition methods of face gender | |
| CN105956570B (en) | Smile recognition method based on lip features and deep learning | |
| CN111401343B (en) | Method for identifying attributes of people in image and training method and device for identification model | |
| CN103839033A (en) | Face identification method based on fuzzy rule | |
| CN109583452B (en) | Human identity identification method and system based on barefoot footprints | |
| CN110969073B (en) | A Facial Expression Recognition Method Based on Feature Fusion and BP Neural Network | |
| CN105320948A (en) | Image based gender identification method, apparatus and system | |
| CN109255339B (en) | Classification method based on self-adaptive deep forest human gait energy map |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: Unit 801, Unit 5, No. 2 Building, 9 Yuan, Fenghao East Road, Haidian District, Beijing 100094 Applicant after: Wen'an Beijing intelligent technology Limited by Share Ltd Address before: 100085 4th Floor, No. 1 Courtyard, Shangdi East Road, Haidian District, Beijing Applicant before: Wen'an Beijing intelligent technology Limited by Share Ltd |
|
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170322 |