CN114694266A - Silent in-vivo detection method, system, equipment and storage medium - Google Patents
Silent in-vivo detection method, system, equipment and storage medium Download PDFInfo
- Publication number
- CN114694266A CN114694266A CN202210316884.6A CN202210316884A CN114694266A CN 114694266 A CN114694266 A CN 114694266A CN 202210316884 A CN202210316884 A CN 202210316884A CN 114694266 A CN114694266 A CN 114694266A
- Authority
- CN
- China
- Prior art keywords
- image
- resolution
- living body
- face
- vector
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 156
- 238000001727 in vivo Methods 0.000 title claims abstract description 57
- 238000000034 method Methods 0.000 claims abstract description 23
- 238000012795 verification Methods 0.000 claims abstract description 7
- 238000012545 processing Methods 0.000 claims description 35
- 230000006870 function Effects 0.000 claims description 29
- 238000004364 calculation method Methods 0.000 claims description 15
- 238000012549 training Methods 0.000 claims description 11
- 238000004590 computer program Methods 0.000 claims description 9
- 238000001574 biopsy Methods 0.000 claims description 8
- 238000012952 Resampling Methods 0.000 claims description 6
- 238000003062 neural network model Methods 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 5
- 230000000694 effects Effects 0.000 abstract description 5
- 238000007796 conventional method Methods 0.000 abstract description 3
- 238000005516 engineering process Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000011478 gradient descent method Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformation in the plane of the image
- G06T3/40—Scaling the whole image or part thereof
- G06T3/4053—Super resolution, i.e. output image resolution higher than sensor resolution
Abstract
The invention discloses a silent in-vivo detection method, a silent in-vivo detection system, silent in-vivo detection equipment and a storage medium, wherein the silent in-vivo detection method comprises the following steps: acquiring a face image; inputting the face image into a first living body detection model to output a first vector; inputting the face image into a super-resolution network to obtain a face super-resolution image, and inputting the face super-resolution image into a second living body detection model to output a second vector; obtaining a live body detection probability based on the first vector and the second vector to output a live body detection result. The invention designs a two-way in-vivo detection model, wherein one way of the model utilizes face super-resolution to assist in-vivo detection, the other way of the model adopts a conventional method to carry out in-vivo detection, and the two ways of detection results are combined to output the final in-vivo detection result. The method not only keeps the original living body detection effect and avoids the occurrence of errors in living body judgment, but also utilizes the face super-resolution network assistance to enable fuzzy photos to become clearer, thereby improving the accuracy of the living body detection and the identity verification.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to a silence living body detection method, a silence living body detection system, silence living body detection equipment and a storage medium.
Background
With the development of artificial intelligence technology, living body detection technology appears, and the living body detection technology is widely applied to bank remote business, face payment and access control systems.
The living body detection technology is mainly used for verifying whether a user operates the living body per se, so that attacking means such as photos, face changing synthesis, masks, head models and the like are resisted, the identity of the user is prevented from being stolen, fraudulent behaviors are screened, and the benefit of the user is ensured.
In the existing in-vivo detection technology, it is generally required to ensure that a face image acquired by a camera is clear, if the in-vivo detection is performed in a poor environment, for example, when light is not good, the imaging of the camera will be blurred, so that the face recognition after the in-vivo detection and the in-vivo detection are successful is greatly affected, and the face recognition ratio is greatly reduced.
Disclosure of Invention
In order to overcome the defects of the prior art, one of the objectives of the present invention is to provide a silent biopsy method, which can perform biopsy on a blurred picture and improve the accuracy of the biopsy.
It is another object of the present invention to provide a silent biopsy system.
It is a further object of the present invention to provide an electronic device.
It is a further object of the present invention to provide a computer readable storage medium.
One of the purposes of the invention is realized by adopting the following technical scheme:
a silent liveness detection method, comprising:
acquiring a face image;
inputting the face image into a first living body detection model to output a first vector;
inputting the face image into a super-resolution network to obtain a face super-resolution image, and inputting the face super-resolution image into a second living body detection model to output a second vector;
obtaining a live body detection probability based on the first vector and the second vector to output a live body detection result.
Further, the method for acquiring the face image comprises the following steps:
and acquiring the video information or picture information obtained by shooting, carrying out face detection on the video information or picture information, and marking the frame image or picture containing the face as a face image.
Further, the method for establishing the first in-vivo detection model comprises the following steps:
acquiring a sample image, and resampling the sample image to obtain an image with a first resolution;
performing image processing on the image with the first resolution, and calculating data obtained by the image processing by using a cross entropy Loss function to obtain an L1Loss function; wherein the image processing comprises convolution processing and linear processing;
training a neural network model based on the L1Loss function to obtain a first in vivo detection model.
Further, the second in-vivo detection model is established by the method comprising the following steps:
resampling the sample image to obtain an image of a second resolution, the second resolution being higher than the first resolution;
performing a super-resolution calculation on the image of the second resolution using the super-resolution network to obtain a super-resolution image;
performing image processing on the super-resolution image, and calculating data obtained by the image processing by using a cross entropy Loss function to obtain an L2Loss function; wherein the image processing comprises convolution processing and linear processing;
training a neural network model based on the L2Loss function to obtain a second in vivo detection model.
Further, the first vector and the second vector are each a vector having three elements; and the sum of the three elements is 1;
wherein, the first element is the probability that the result of the living body detection is a real living body; the second element is the probability that the result of the in vivo test is a photograph; the third element is the probability that the biopsy result is a mask pattern.
Further, the method for calculating and obtaining the living body detection probability comprises the following steps:
the live body detection probability is the first vector first coefficient + the second vector second coefficient.
Further, the method for outputting the result of the in-vivo test comprises the following steps:
judging whether the living body detection probability is greater than a preset value, if so, outputting a living body detection result, and outputting the face super-resolution image to carry out face identification verification on the face super-resolution image; otherwise, outputting the detection result of the non-living body.
The second purpose of the invention is realized by adopting the following technical scheme:
a silent liveness detection system performing the silent liveness detection method as described above, the system comprising:
the image recognition module is used for acquiring a face image;
a first living body detection model for calculating the face image to output a first vector;
the super-resolution network model is used for processing the face image to output a face super-resolution image;
the second living body detection model is used for processing the face super-resolution image to output a second vector;
and the living body detection module is used for calculating and obtaining a living body detection probability based on the first vector and the second vector so as to output a living body detection result.
The third purpose of the invention is realized by adopting the following technical scheme:
an electronic device comprising a processor, a memory and a computer program stored on the memory and executable on the processor, the processor implementing the silent liveness detection method as described above when executing the computer program.
The fourth purpose of the invention is realized by adopting the following technical scheme:
a computer-readable storage medium, having stored thereon a computer program which, when executed, implements a silent liveness detection method as described above.
Compared with the prior art, the invention has the beneficial effects that:
the invention designs a double-path living body detection model, wherein one path of the living body detection model utilizes the face super-resolution to assist the living body detection, the other path of the living body detection model adopts a conventional method to carry out the living body detection, and the final living body detection result is output by combining the detection results of the two paths of the living body detection model. The method not only keeps the original living body detection effect and avoids the occurrence of errors in living body judgment, but also utilizes the face super-resolution network assistance to enable fuzzy photos to become clearer, thereby improving the accuracy of the living body detection and the identity verification.
Drawings
FIG. 1 is a schematic diagram of a first in-vivo detection model construction process according to the present invention;
FIG. 2 is a schematic diagram of a second in-vivo detection model according to the present invention;
FIG. 3 is a schematic flow chart of the in-vivo detection method of the present invention;
FIG. 4 is a block diagram of the in-vivo detection system of the present invention.
Detailed Description
The present invention will be further described with reference to the accompanying drawings and the detailed description, and it should be noted that any combination of the embodiments or technical features described below can be used to form a new embodiment without conflict.
Example one
The embodiment provides a silent in-vivo detection method, which is used for solving the problem that a living body cannot be normally identified due to fuzzy pictures or identity authentication cannot be performed by fuzzy pictures after the living body is detected, and improving the accuracy of in-vivo detection and identity authentication.
Before the in-vivo detection, a first in-vivo detection model and a second in-vivo detection model need to be established in advance, wherein the first in-vivo detection model and the second in-vivo detection model are both used for in-vivo detection of the face in the image, but the first in-vivo detection model adopts a conventional method for in-vivo detection, so that the original in-vivo detection effect is reserved; and the second living body detection model utilizes the face super-resolution to assist in living body detection, so that a blurred picture can be clearer, and living body detection accuracy is improved by carrying out living body detection on the clear picture.
Referring to fig. 1 and fig. 2, the first living body detection model and the second living body detection model are established by the following method:
step S1: acquiring a sample image, wherein the sample image is a picture containing a human face; resampling the sample image to obtain an image of a first resolution and an image of a second resolution;
wherein the second resolution is higher than the first resolution; the image of the first resolution is used to train the first in vivo detection model; the image of the second resolution is used for training the second in vivo detection model; in this embodiment, the first resolution is 80 × 80; the second resolution is 512 x 512.
Step S2: performing image processing on the image with the first resolution, and calculating data obtained by the image processing by using a cross entropy Loss function to obtain an L1Loss function;
the image processing includes a plurality of convolution calculations and a plurality of linear calculations, for example:
step S21: performing convolution block calculation on the image with the resolution of 80 by 80 to obtain data of 40 by 32;
step S22: performing a volume block calculation on the 40 × 32 data to obtain 20 × 64 data;
step S23: performing a depth separable volume block calculation on the 20 × 64 data to obtain 10 × 128 data;
step S24: performing linear block computation on 10 × 128 data to obtain 1 × 128 data;
step S25: performing linear calculation on the data of 1 × 128 to obtain data of 1 × 3;
after the image is subjected to feature extraction through the image processing, model training is carried out by utilizing data after the image processing, and the method comprises the following steps:
step S26: calculating 1 × 3 data using a cross entropy Loss function to obtain an L1Loss function;
the cross entropy loss function is generally used for measuring the difference value of two probability distributions of the calculation probability and the real probability, and the cross entropy loss function is used for measuring the error degree of the in-vivo detection in the embodiment; the L1Loss function is also referred to as a minimum absolute value deviation Loss function, and represents an error obtained by making an absolute value of a target value (true value) and a model output (estimated value).
Step S3: training a neural network model based on the L1Loss function to obtain a first in vivo detection model; and in the process of training by using the L1Loss function to obtain the first in-vivo detection model, optimizing a neural network by adopting an SGD gradient descent method.
Step S4: performing super-resolution calculation on the image of the second resolution obtained by resampling in step S1, that is, calculating an image of 512 × 512 resolution using a super-resolution network model of the human face to obtain a super-resolution image. Among them, the method for establishing the super-resolution network model of the human face is disclosed in the prior art and is not described in detail here.
Step S5: performing image processing on the super-resolution image, and calculating data obtained by the image processing by using a cross entropy Loss function to obtain an L2Loss function;
the method for processing the super-resolution image comprises the following steps:
step S51: performing convolution block calculation on the super-resolution image to obtain 64 × 128 data;
step S52: performing a depth separable convolution block calculation on 64 x 128 data to obtain 8 x 128 data;
step S53: performing linear block calculations on 8 x 128 data to obtain 1 x 128 data;
step S54: linearizing the data of 1 × 128 to obtain data of 1 × 3;
step S55: calculating 1 x 3 data by using a cross entropy Loss function to obtain an L2Loss function; the L2Loss function is also referred to as a least mean square deviation Loss function, and represents an error obtained by subtracting a target value (true value) from a model output (estimated value) and then squaring the difference.
Step S6: training a neural network model based on the L2Loss function to obtain a second in vivo detection model; the optimization method also adopts an SGD gradient descent method for optimization.
In the embodiment, the fuzzy picture is subjected to super-resolution processing, and then the super-resolution image of the face is used for second living body detection model training, so that the accuracy of real human living body detection can be improved, and the situation of large detection error caused by fuzzy pictures is avoided.
After the first and second in-vivo detection models are constructed by the model training method of the above steps S1-S6, the in-vivo detection prediction can be performed on the images taken in real time, as shown in fig. 3, the in-vivo detection method includes the following steps:
step A: acquiring a face image;
the face image can be obtained from a video stream shot in real time or a picture shot in real time, namely, the obtained video information or picture information is subjected to face detection, and a frame image containing a face in the video is intercepted to be used as the face image; or screening the shot pictures, marking the pictures containing the human faces as human face images, and removing the pictures not containing the human faces. And if the face is not identified in the video or the picture, continuously acquiring the video and picture information until a face image containing the face is obtained.
And B: inputting the face image into the first living body detection model trained in advance, and calculating to obtain a first vector V1;
the first vector V1 is a vector having three elements; and the sum of the three elements is 1;
wherein, the first element of the first vector V1 is the probability that the result of the living body detection is a real living body; the second element is the probability that the result of the in vivo test is a photograph; the third element is the probability that the biopsy result is a mask pattern.
Step C: inputting the face image into a super-resolution network to obtain a face super-resolution image, and inputting the face super-resolution image into a second living body detection model to output a second vector V2;
the second vector V2 is likewise a three-element vector; and the sum of the three elements is also 1;
wherein, the first element of the second vector V2 is the probability that the result of the living body detection is a real living body; the second element is the probability that the result of the in vivo test is a photograph; the third element is the probability that the biopsy result is a mask pattern.
Step S4: obtaining a live body detection probability based on the first vector and the second vector to output a live body detection result.
The method for calculating the living body detection probability comprises the following steps:
the living body detection probability V ═ the first vector V1 × -the first coefficient + the second vector V2 × -the second coefficient. The specific values of the first coefficient and the second coefficient may be adjusted according to actual conditions, in this embodiment, the first coefficient is 0.2, and the second coefficient is 0.8, so as to calculate the final probability.
Then, judging whether the living body detection probability is greater than a preset value, and if the living body detection probability is greater than the preset value, outputting a living body detection result; otherwise, outputting a detection result of the non-living body; in this embodiment, the preset value is 0.5, and when the living body detection probability is greater than 0.5, it can be determined as a living body.
In addition, the living body detection can be applied to the field of identity verification of face identification, namely when the living body detection result is output as a 'living body', the face super-resolution image obtained in the step C is output to an identity verification module, and the face identification is carried out on the super-resolution image so as to realize identity verification; and if the living body detection result is output as 'non-living body', the identity authentication is not passed.
In the embodiment, the two-path in-vivo detection method is adopted, so that the original in-vivo detection effect is maintained, the non-living body is prevented from being judged as the living body, in addition, the face super-resolution network assistance is utilized, the fuzzy photos can be clearer, the effect on the in-vivo detection of the real person is better, the obtained clear image is used for subsequent face comparison and identification, and the identification rate of the face comparison and identification can be improved.
Example two
The present embodiment provides a silent liveness detection system, which executes the silent liveness detection method according to the first embodiment, as shown in fig. 4, the system includes:
the image recognition module is used for acquiring a face image;
a first living body detection model for calculating the face image to output a first vector;
the super-resolution network model is used for processing the face image to output a face super-resolution image;
the second living body detection model is used for processing the face super-resolution image to output a second vector;
and the living body detection module is used for obtaining a living body detection probability based on the first vector and the second vector calculation so as to output a living body detection result.
Furthermore, in some embodiments, an electronic device is provided, which comprises a processor, a memory, and a computer program stored on the memory and executable on the processor, the processor implementing the silence liveness detection method of embodiment one when executing the computer program; in addition, a computer-readable storage medium is also provided in some embodiments, having stored thereon a computer program that, when executed, implements the silent liveness detection method described above.
The system, the device, and the storage medium in this embodiment are based on multiple aspects of the same inventive concept, and the method implementation process has been described in detail in the foregoing, so that those skilled in the art can clearly understand the structure and implementation process of the system, the device, and the storage medium in this embodiment according to the foregoing description, and for the sake of brevity of the description, no further description is given here.
The above embodiments are only preferred embodiments of the present invention, and the protection scope of the present invention is not limited thereby, and any insubstantial changes and substitutions made by those skilled in the art based on the present invention are within the protection scope of the present invention.
Claims (10)
1. A silent liveness detection method, comprising:
acquiring a face image;
inputting the face image into a first living body detection model to output a first vector;
inputting the face image into a super-resolution network to obtain a face super-resolution image, and inputting the face super-resolution image into a second living body detection model to output a second vector;
obtaining a live body detection probability based on the first vector and the second vector to output a live body detection result.
2. The silent in-vivo detection method as claimed in claim 1, wherein the method for acquiring the face image is as follows:
and acquiring the video information or picture information obtained by shooting, carrying out face detection on the video information or picture information, and marking the frame image or picture containing the face as a face image.
3. The silent in-vivo detection method as claimed in claim 1, wherein the first in-vivo detection model is established by:
acquiring a sample image, and resampling the sample image to obtain an image with a first resolution;
performing image processing on the image with the first resolution, and calculating data obtained by the image processing by using a cross entropy Loss function to obtain an L1Loss function; wherein the image processing comprises convolution processing and linear processing;
training a neural network model based on the L1Loss function to obtain a first in vivo detection model.
4. The silent in-vivo detection method as claimed in claim 3, wherein the second in-vivo detection model is established by:
resampling the sample image to obtain an image of a second resolution, the second resolution being higher than the first resolution;
performing a super-resolution calculation on the image of the second resolution using the super-resolution network to obtain a super-resolution image;
performing image processing on the super-resolution image, and calculating data obtained by the image processing by using a cross entropy Loss function to obtain an L2Loss function; wherein the image processing comprises convolution processing and linear processing;
and training a neural network model based on the L2Loss function to obtain a second in-vivo detection model.
5. The silent liveness detection method of claim 1, wherein the first vector and the second vector are each a vector having three elements; and the sum of the three elements is 1;
wherein, the first element is the probability that the living body detection result is a real living body; the second element is the probability that the result of the in vivo test is a photograph; the third element is the probability that the biopsy result is a mask pattern.
6. The silent liveness detection method according to claim 5, wherein the method of calculating the probability of obtaining the liveness detection is:
the live body detection probability is the first vector first coefficient + the second vector second coefficient.
7. The silent liveness detection method according to claim 1, wherein the method of outputting the liveness detection result is:
judging whether the living body detection probability is larger than a preset value, if so, outputting a living body detection result and outputting the face super-resolution image to carry out face recognition verification on the face super-resolution image; otherwise, outputting the detection result of the non-living body.
8. A silent liveness detection system, characterized in that it performs the silent liveness detection method according to any one of claims 1 to 7, said system comprising:
the image recognition module is used for acquiring a face image;
a first living body detection model for calculating the face image to output a first vector;
the super-resolution network model is used for processing the face image to output a face super-resolution image;
the second living body detection model is used for processing the face super-resolution image to output a second vector;
and the living body detection module is used for obtaining a living body detection probability based on the first vector and the second vector calculation so as to output a living body detection result.
9. An electronic device, comprising a processor, a memory, and a computer program stored on the memory and executable on the processor, wherein the processor implements the silent liveness detection method of any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, having stored thereon a computer program which, when executed, implements the silent liveness detection method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210316884.6A CN114694266A (en) | 2022-03-28 | 2022-03-28 | Silent in-vivo detection method, system, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210316884.6A CN114694266A (en) | 2022-03-28 | 2022-03-28 | Silent in-vivo detection method, system, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114694266A true CN114694266A (en) | 2022-07-01 |
Family
ID=82140229
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210316884.6A Pending CN114694266A (en) | 2022-03-28 | 2022-03-28 | Silent in-vivo detection method, system, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114694266A (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105389554A (en) * | 2015-11-06 | 2016-03-09 | 北京汉王智远科技有限公司 | Face-identification-based living body determination method and equipment |
| US9361667B2 (en) * | 2010-08-27 | 2016-06-07 | Sony Corporation | Image processing apparatus and method |
| CN108549886A (en) * | 2018-06-29 | 2018-09-18 | 汉王科技股份有限公司 | A kind of human face in-vivo detection method and device |
| US20200005061A1 (en) * | 2018-06-28 | 2020-01-02 | Beijing Kuangshi Technology Co., Ltd. | Living body detection method and system, computer-readable storage medium |
| WO2020078243A1 (en) * | 2018-10-19 | 2020-04-23 | 阿里巴巴集团控股有限公司 | Image processing and face image identification method, apparatus and device |
| CN111931594A (en) * | 2020-07-16 | 2020-11-13 | 广州广电卓识智能科技有限公司 | Face recognition living body detection method and device, computer equipment and storage medium |
| CN112784857A (en) * | 2021-01-29 | 2021-05-11 | 北京三快在线科技有限公司 | Model training and image processing method and device |
| CN113065579A (en) * | 2021-03-12 | 2021-07-02 | 支付宝(杭州)信息技术有限公司 | Method and device for classifying target object |
-
2022
- 2022-03-28 CN CN202210316884.6A patent/CN114694266A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9361667B2 (en) * | 2010-08-27 | 2016-06-07 | Sony Corporation | Image processing apparatus and method |
| CN105389554A (en) * | 2015-11-06 | 2016-03-09 | 北京汉王智远科技有限公司 | Face-identification-based living body determination method and equipment |
| US20200005061A1 (en) * | 2018-06-28 | 2020-01-02 | Beijing Kuangshi Technology Co., Ltd. | Living body detection method and system, computer-readable storage medium |
| CN108549886A (en) * | 2018-06-29 | 2018-09-18 | 汉王科技股份有限公司 | A kind of human face in-vivo detection method and device |
| WO2020078243A1 (en) * | 2018-10-19 | 2020-04-23 | 阿里巴巴集团控股有限公司 | Image processing and face image identification method, apparatus and device |
| CN111931594A (en) * | 2020-07-16 | 2020-11-13 | 广州广电卓识智能科技有限公司 | Face recognition living body detection method and device, computer equipment and storage medium |
| CN112784857A (en) * | 2021-01-29 | 2021-05-11 | 北京三快在线科技有限公司 | Model training and image processing method and device |
| CN113065579A (en) * | 2021-03-12 | 2021-07-02 | 支付宝(杭州)信息技术有限公司 | Method and device for classifying target object |
Non-Patent Citations (2)
| Title |
|---|
| MEIGUI ZHANG ET AL.: "A Survey on Face Anti-Spoofing Algorithms", 《JOURNAL OF INFORMATION HIDING AND PRIVACY PROTECTION》 * |
| 卢子谦等: "人脸反欺诈活体检测综述", 《信息安全学报》, vol. 5, no. 2 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113196289B (en) | Human body action recognition method, human body action recognition system and equipment | |
| JP5459674B2 (en) | Moving object tracking system and moving object tracking method | |
| CN107944395B (en) | Method and system for verifying and authenticating integration based on neural network | |
| CN110378219B (en) | Living body detection method, living body detection device, electronic equipment and readable storage medium | |
| CN112991345B (en) | Image authenticity detection method and device, computer equipment and storage medium | |
| CN107767358B (en) | Method and device for determining ambiguity of object in image | |
| CN110705353A (en) | Method and device for identifying face to be shielded based on attention mechanism | |
| CN116403294B (en) | Transformer-based multi-view width learning living body detection method, medium and equipment | |
| CN111414858A (en) | Face recognition method, target image determination method, device and electronic system | |
| CN113705426A (en) | Face checking method, device, server and readable storage medium | |
| CN116453232A (en) | Face living body detection method, training method and device of face living body detection model | |
| CN114037838A (en) | Neural network training method, electronic device and computer program product | |
| CN111784658A (en) | Quality analysis method and system for face image | |
| CN113642639B (en) | Living body detection method, living body detection device, living body detection equipment and storage medium | |
| CN109325472B (en) | Face living body detection method based on depth information | |
| CN112949453B (en) | Training method of smoke and fire detection model, smoke and fire detection method and equipment | |
| CN111696090B (en) | Method for evaluating quality of face image in unconstrained environment | |
| CN111046755A (en) | Character recognition method, character recognition device, computer equipment and computer-readable storage medium | |
| CN114694266A (en) | Silent in-vivo detection method, system, equipment and storage medium | |
| CN116206373A (en) | Living body detection method, electronic device and storage medium | |
| CN114332446B (en) | Image countermeasure sample generation method with rotation robustness in physical world | |
| CN116152542A (en) | Training method, device, equipment and storage medium for image classification model | |
| KR100567765B1 (en) | System and Method for face recognition using light and preprocess | |
| CN115273245A (en) | Living body detection method, living body detection device and computer-readable storage medium | |
| CN113554685A (en) | Method and device for detecting moving target of remote sensing satellite, electronic equipment and storage medium |
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 | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Country or region after: China Address after: No. 001-030, Yuntong space, office building, No. 9, Kelin Road, Science City, Guangzhou hi tech Industrial Development Zone, Guangzhou, Guangdong 510000 Applicant after: GRG TALLY-VISION I.T. Co.,Ltd. Applicant after: Guangdian Yuntong Group Co.,Ltd. Address before: No. 001-030, Yuntong space, office building, No. 9, Kelin Road, Science City, Guangzhou hi tech Industrial Development Zone, Guangzhou, Guangdong 510000 Applicant before: GRG TALLY-VISION I.T. Co.,Ltd. Country or region before: China Applicant before: GRG BANKING EQUIPMENT Co.,Ltd. |
|
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20240123 Address after: No. 001-030, Yuntong space, office building, No. 9, Kelin Road, Science City, Guangzhou hi tech Industrial Development Zone, Guangzhou, Guangdong 510000 Applicant after: GRG TALLY-VISION I.T. Co.,Ltd. Country or region after: China Address before: No. 001-030, Yuntong space, office building, No. 9, Kelin Road, Science City, Guangzhou hi tech Industrial Development Zone, Guangzhou, Guangdong 510000 Applicant before: GRG TALLY-VISION I.T. Co.,Ltd. Country or region before: China Applicant before: Guangdian Yuntong Group Co.,Ltd. |