CN113361395B - AI face-changing video detection method based on multitask learning model - Google Patents
AI face-changing video detection method based on multitask learning model Download PDFInfo
- Publication number
- CN113361395B CN113361395B CN202110624171.1A CN202110624171A CN113361395B CN 113361395 B CN113361395 B CN 113361395B CN 202110624171 A CN202110624171 A CN 202110624171A CN 113361395 B CN113361395 B CN 113361395B
- Authority
- CN
- China
- Prior art keywords
- video
- detection
- user
- face
- data
- 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.)
- Expired - Fee Related
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 72
- 241000700605 Viruses Species 0.000 claims abstract description 16
- 238000012216 screening Methods 0.000 claims abstract description 11
- 238000006243 chemical reaction Methods 0.000 claims description 14
- 238000012163 sequencing technique Methods 0.000 claims description 13
- 238000004040 coloring Methods 0.000 claims description 10
- 238000000034 method Methods 0.000 claims description 9
- 238000007639 printing Methods 0.000 claims description 9
- 230000003993 interaction Effects 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 4
- 238000013481 data capture Methods 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000012549 training Methods 0.000 claims description 3
- 238000013135 deep learning Methods 0.000 abstract description 3
- 208000015181 infectious disease Diseases 0.000 abstract 1
- 238000005516 engineering process Methods 0.000 description 7
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000008921 facial expression Effects 0.000 description 1
- 238000011895 specific detection Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Linguistics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an AI face-changing video detection method based on a multitask learning model, belonging to the field of computer vision and deep learning, and comprising the following specific steps: (1) uploading a video to be detected by a user; (2) constructing a face image forgery detector; (3) carrying out frame-by-frame image taking on the video; (4) screening the extracted pictures; (5) carrying out face changing detection on the rest pictures; (6) feeding back the detection result to the user; the face image counterfeiting detector can avoid computer infection with viruses, reduce the risk of stealing user information, protect the property safety of users, meanwhile, the face image counterfeiting detector can continuously update and learn, continuously improve the working efficiency of the face image counterfeiting detector, facilitate checking of workers, prevent the manual recording of the workers from generating errors, and improve the working efficiency of the workers.
Description
Technical Field
The invention relates to the field of computer vision and deep learning, in particular to an AI face changing video detection method based on a multitask learning model.
Background
The AI face changing is to change the face of another person into the face of the other person through an AI artificial intelligence technology, the face expression is natural, the effect is vivid, only one photo needs to be taken in the whole process, then the AI technology can be used for replacing people in a drama or a small video on ZAO software, so that a video with the person as a main role is generated, along with the development of deep learning, the AI face changing technology is more and more mature, the AI face changing effect is better and more, along with the development of the AI face changing technology, a lot of negative influences are brought, for example, deepFakes show how to use computer graphics and visual technologies to change faces in video, further the reputation of the other person is damaged, along with the continuous development of science and technology, the fidelity of the AI face changing technology is gradually improved, and sometimes, the authenticity of a human cannot be distinguished; therefore, it becomes more important to invent an AI face-changing video detection method based on a multitask learning model;
through retrieval, chinese patent No. CN111950497A discloses an AI face-changing video detection method based on a multitask learning model, although the calculation cost is reduced and the accuracy is improved, when a network data set is collected, the security detection is not carried out on the network data set, a computer is infected with viruses, user information is easily stolen, and the property safety of the user is harmed; in addition, the conventional AI face-changing video detection method based on the multitask learning model cannot effectively feed back the detection result to the user, so that the user recording error is easily caused, and the working quality of the user is reduced; therefore, an AI face-changing video detection method based on a multitask learning model is provided.
Disclosure of Invention
The invention aims to solve the defects in the prior art and provides an AI face changing video detection method based on a multitask learning model.
In order to achieve the purpose, the invention adopts the following technical scheme:
an AI face-changing video detection method based on a multitask learning model comprises the following specific steps:
(1) Uploading a video to be detected by a user: a user uploads a video to be detected to a computer through an external input device, wherein the external input device is one of a keyboard, a mouse or a touch screen, the video is successfully uploaded, and the user puts the video into detection software;
(2) Constructing a face image counterfeiting detector: the detection software receives the video, starts to perform data interaction with the Internet, performs data capture on face-changing detection data, performs security detection on the captured data, and starts to construct a face image counterfeiting detector;
(3) Taking pictures of the video frame by frame: the face image counterfeiting detector starts to automatically scan the video to be detected, calculates the video quantity, and simultaneously performs detection sequencing, the video sequencing is completed, and the face image counterfeiting detector performs frame-by-frame image taking on the video;
(4) And (3) carrying out picture screening on the extracted pictures: after the picture is extracted, starting to construct a picture screener, wherein the picture screener starts to perform data interaction with the characteristic information base and starts to perform contrast screening on the extracted picture;
(5) Face changing detection is carried out on the rest pictures: performing forgery detection on the screened pictures, respectively recording detection results into an XLSX worksheet, and marking detection time in the XLSX worksheet;
(6) And feeding back the detection result to a user: and feeding back the detection result to a user through display equipment, searching and checking an XLSX worksheet through input equipment of the user, and printing the XLSX worksheet through printing equipment, wherein the display equipment is a CRT display screen, an LCD display screen or an LED display screen, and the printing equipment is a laser printer, an ink-jet printer or a stylus printer.
Further, the safety detection in the step (2) specifically comprises the following steps:
the method comprises the following steps: capturing FaceForensics, face2Face, faceSwap and DeepFakes data sets in real time from the Internet, and starting a firewall;
step two: the firewall starts to detect the data in each data set, deletes the virus data in the data set, and records the deleted virus data in the cloud virus information base;
step three: and after the data detection is finished, training a faceforces, a Face2Face, a FaceSwap and a DeepFakes data set into a current optimal human Face image counterfeiting detector.
Further, the detection sorting in step (3) specifically comprises the following steps:
the first step is as follows: starting to collect video information uploaded by a user, and sequencing the video information according to the uploading time;
the second step is that: the user arranges and adjusts the video through the external input equipment, and the human face image counterfeiting detector starts to update the video arrangement sequence according to the user adjustment information.
Further, the specific step of taking pictures frame by frame in step (3) is as follows:
s1: performing video segmentation processing on a video to be detected according to a frame, and respectively marking segments which are segmented as A1, A2, A3, 8230An, wherein n is a natural number and the size of n is increased in sequence;
s2: and converting the A1-An into pictures and sequencing the pictures according to the cutting time.
5. The AI face-changing video detection method based on the multitask learning model according to claim 1, further comprising the specific steps of comparing and screening in the step (4) as follows:
and (4) SS1: the picture filter begins to extract human characteristic information from the characteristic information base and processes the human characteristic information to generate comparison data;
and SS2: respectively comparing A1-An with the comparison data, deleting pictures which do not contain human characteristics, and respectively marking the screened pictures as B1, B2, B3, \8230andBm, wherein m is a natural number, and the sizes of m are sequentially increased.
Further, the counterfeit detection in step (5) specifically comprises the following steps:
p1: sequentially carrying out color space conversion on the B1-Bm, and starting to compare the difference of the front edge, the back edge, the texture and the surface of the conversion;
p2: if the low contrast edge and the high contrast edge are almost as bright, texture difference starts to be contrasted, and if the brightness of the low contrast edge and the brightness of the high contrast edge are obviously different, the video is judged to be a face changing video;
p3: if the coloring of the area with more texture details is higher than the smooth surface, the surface difference is compared, and if the coloring of the area with more texture details is lower than the smooth surface, the video is judged to be a face changing video;
p4: if the surfaces are colored uniformly before and after conversion, judging that the video is not a face change video, and if the surfaces are colored non-uniformly before and after conversion, judging that the video is a face change video;
p5: and after the detection is finished, performing data matching on the video name and the judgment result, simultaneously orderly inputting the matched data into an XLSX worksheet, and marking the detection time in the XLSX worksheet.
Further, the specific steps of searching and checking in the step (6) are as follows:
PP1: inputting a primary time period X to be searched by a user through input equipment;
and (3) PP2: inputting the secondary time period X again after the user inputs the primary time period X;
and (3) PP: after the user inputs the primary time period X and the secondary time period X, the computer calls the content required by the user and displays the content through the display device.
Compared with the prior art, the invention has the beneficial effects that:
1. the AI Face-changing video detection method based on the multitask learning model receives a video uploaded by a user, starts data interaction with the Internet, simultaneously starts to capture FaceForensics, face2Face, face swap and DeepFakes data sets from the Internet in real time, performs virus detection on the captured data sets through a firewall, deletes virus data in the captured data sets, records the deleted virus data in a cloud virus information base, completes data detection, trains the data sets into a current optimal Face image counterfeiting detector, and updates the Face image counterfeiting detector in real time, so that the computer can be prevented from infecting viruses, the risk that user information is stolen is reduced, the property safety of the user is protected, and meanwhile, the Face image detector can continuously perform updating and counterfeiting learning, and the working efficiency of the Face image counterfeiting detector is continuously improved;
2. according to the AI face-changing video detection method based on the multitask learning model, a video to be detected is cut into a plurality of groups of pictures, the pictures are screened through a picture screening device, color space conversion processing is carried out on the screened pictures, edge brightness difference, texture tinting strength difference and the tinting strength before and after surface conversion are compared, data matching is carried out on video names and judgment results, meanwhile, the matched data are orderly recorded into an XLSX worksheet, detection time is marked in the XLSX worksheet, the detection results can be orderly recorded, the detection method is convenient for workers to check, the occurrence of errors in manual recording of the workers is prevented, and the working efficiency of the workers is improved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention.
Fig. 1 is a flow chart of an AI face-changing video detection method based on a multitask learning model according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments.
In the description of the present invention, it is to be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", and the like, indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, are merely for convenience in describing the present invention and simplifying the description, and do not indicate or imply that the device or element being referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus, should not be construed as limiting the present invention.
Referring to fig. 1, the embodiment discloses an AI face-changing video detection method based on a multitask learning model, which specifically includes the following steps:
(1) Uploading a video to be detected by a user: a user uploads a video to be detected to a computer through external input equipment, the video is successfully uploaded, and the user puts the video into detection software;
specifically, the external input device is one of a keyboard, a mouse, or a touch screen.
(2) Constructing a face image counterfeiting detector: the detection software receives the video, starts to perform data interaction with the Internet, performs data capture on face-changing detection data, performs security detection on the captured data, and starts to construct a face image counterfeiting detector.
The embodiment provides a safety detection method, which comprises the following specific safety detection steps:
the method comprises the following steps: capturing FaceForensics, face2Face, faceSwap and DeepFakes data sets in real time from the Internet, and starting a firewall;
step two: the firewall starts to detect the data in each data set, deletes the virus data in the data set, and records the deleted virus data in the cloud virus information base;
step three: and (3) finishing data detection, and training a faceForensics, face2Face, faceSwap and DeepFakes data set into a current optimal Face image counterfeiting detector.
(3) Taking pictures of the video frame by frame: the face image counterfeiting detector starts to scan the video to be detected, calculates the video quantity, performs detection sequencing at the same time, finishes video sequencing, and performs frame-by-frame image extraction on the video.
The embodiment provides a detection sorting method, which comprises the following specific detection sorting steps:
the first step is as follows: starting to collect video information uploaded by a user, and sequencing the video information according to the uploading time;
the second step is that: the user arranges and adjusts the video through the external input equipment, and the human face image counterfeiting detector starts to update the video arrangement sequence according to the user adjustment information.
The embodiment further discloses a frame-by-frame image-taking method, which comprises the following specific steps:
s1: performing video segmentation processing on a video to be detected according to a frame, and respectively marking segmented segments as A1, A2, A3, 8230An, wherein n is a natural number and the size of n is increased in sequence;
s2: and converting the A1-An into pictures, and sequencing the pictures according to the cutting time.
(4) And (3) carrying out picture screening on the extracted pictures: and after the picture is extracted, starting to construct a picture screener, completing the construction, starting to perform data interaction with the characteristic information base by the picture screener, and starting to perform contrast screening on the extracted picture.
Specifically, the steps of the comparison and screening of the image screener are as follows:
SS1: the picture filter begins to extract human characteristic information from the characteristic information base and processes the human characteristic information to generate comparison data;
and SS2: comparing A1-An with the comparison data respectively, deleting pictures which do not contain human characteristics, and marking the screened pictures as B1, B2, B3, \ 8230, bm respectively, wherein m is a natural number, and the size of m is increased sequentially.
(5) Face changing detection is carried out on the rest pictures: and performing forgery detection on the screened pictures, recording detection results into an XLSX worksheet, and marking detection time in the XLSX worksheet.
The embodiment provides a forgery detection method, which specifically comprises the following steps:
p1: sequentially carrying out color space conversion on the B1-Bm, and starting to compare the difference of the front edge, the back edge, the texture and the surface of the conversion;
p2: if the low contrast edge and the high contrast edge are almost as bright, texture difference starts to be contrasted, and if the brightness of the low contrast edge and the brightness of the high contrast edge are obviously different, the video is judged to be a face changing video;
p3: if the coloring of the area with more texture details is higher than the smooth surface, the surface difference is compared, and if the coloring of the area with more texture details is lower than the smooth surface, the video is judged to be a face changing video;
p4: if the coloring of the surface is consistent before and after conversion, judging that the video is not a face changing video, and if the coloring of the surface is inconsistent before and after conversion, judging that the video is a face changing video;
p5: and after the detection is finished, performing data matching on the video name and the judgment result, orderly inputting the matched data into an XLSX working table, and marking detection time in the XLSX working table.
(6) And feeding back the detection result to a user: and feeding back the detection result to a user through a display device, and simultaneously, searching and checking the XLSX worksheet through an input device by the user, and printing the XLSX worksheet through a printing device.
Specifically, the user search and check comprises the following specific steps:
PP1: inputting a primary time period X to be searched by a user through input equipment;
and (3) PP2: inputting the secondary time period X again after the user inputs the primary time period X;
and (3) PP3: after the user inputs the primary time period X and the secondary time period X, the computer calls out the content required by the user and displays the content through the display equipment;
in this embodiment, the display device is specifically a CRT display screen, an LCD display screen, or an LED display screen, and the printing device is specifically a laser printer, an inkjet printer, or a stylus printer.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be considered to be within the technical scope of the present invention, and the technical solutions and the inventive concepts thereof according to the present invention should be equivalent or changed within the scope of the present invention.
Claims (2)
1. An AI face-changing video detection method based on a multitask learning model is characterized by comprising the following specific steps:
(1) Uploading a video to be detected by a user: a user uploads a video to be detected to a computer through an external input device, wherein the external input device is one of a keyboard, a mouse or a touch screen, the video is successfully uploaded, and the user puts the video into detection software;
(2) Constructing a face image counterfeiting detector: the detection software receives the video, starts to perform data interaction with the Internet, performs data capture on face-changing detection data, performs security detection on the captured data, and starts to construct a face image counterfeiting detector;
(3) Taking pictures of the video frame by frame: the face image counterfeiting detector starts to automatically scan the video to be detected, calculates the video quantity, and simultaneously performs detection sequencing, the video sequencing is completed, and the face image counterfeiting detector performs frame-by-frame image taking on the video;
(4) And (3) carrying out picture screening on the extracted pictures: after the picture is extracted, starting to construct a picture screener, wherein the picture screener starts to perform data interaction with the characteristic information base and starts to perform contrast screening on the extracted picture;
(5) Face changing detection is carried out on the rest pictures: performing forgery detection on the screened pictures, respectively recording detection results into an XLSX worksheet, and marking detection time in the XLSX worksheet;
(6) And feeding back the detection result to a user: feeding back the detection result to a user through a display device, and simultaneously searching and checking an XLSX working table through an input device by the user, and printing the XLSX working table through a printing device, wherein the display device is specifically a CRT display screen, an LCD display screen or an LED display screen, and the printing device is specifically a laser printer, an ink-jet printer or a stylus printer;
the safety detection comprises the following specific steps:
the method comprises the following steps: capturing FaceForensics, face2Face, faceSwap and DeepFakes data sets in real time from the Internet, and starting a firewall;
step two: the firewall starts to detect the data in each data set, deletes the virus data in the data set, and records the deleted virus data in the cloud virus information base;
step three: after the data detection is finished, training a faceforces, a Face2Face, a FaceSwap and a DeepFakes data set into a current optimal human Face image counterfeiting detector;
the detection sequencing comprises the following specific steps:
the first step is as follows: starting to collect video information uploaded by a user, and sequencing the video information according to the uploading time;
the second step is that: the user arranges and adjusts the video through the external input equipment, and the human face image counterfeiting detector starts to update the video arrangement sequence according to the user adjustment information;
the specific steps of frame-by-frame image taking are as follows:
s1: performing video segmentation processing on a video to be detected according to a frame, and respectively marking segments which are segmented as A1, A2, A3, 8230An, wherein n is a natural number and the size of n is increased in sequence;
s2: converting A1-An into pictures, and sequencing the pictures according to cutting time;
the specific steps of the comparison screening are as follows:
and (4) SS1: the picture filter begins to extract human characteristic information from the characteristic information base and processes the human characteristic information to generate comparison data;
and (4) SS2: respectively comparing A1-An with the comparison data, deleting pictures which do not contain human characteristics, and respectively marking the screened pictures as B1, B2, B3, \8230, and Bm, wherein m is a natural number, and the sizes of m are sequentially increased;
the counterfeit detection comprises the following specific steps:
p1: sequentially carrying out color space conversion on the B1-Bm, and starting to compare the difference of the front edge, the back edge, the texture and the surface of the conversion;
p2: if the low contrast edge and the high contrast edge are as bright as each other, starting to contrast texture differences, and if the brightness of the low contrast edge and the brightness of the high contrast edge are different, judging that the video is a face changing video;
p3: if the coloring of the area with more texture details is higher than that of the smooth surface, comparing the surface difference, and if the coloring of the area with more texture details is lower than that of the smooth surface, judging that the video is a face changing video;
p4: if the coloring of the surface is consistent before and after conversion, judging that the video is not a face changing video, and if the coloring of the surface is inconsistent before and after conversion, judging that the video is a face changing video;
p5: and after the detection is finished, performing data matching on the video name and the judgment result, orderly inputting the matched data into an XLSX working table, and marking detection time in the XLSX working table.
2. The AI face-changing video detection method based on the multitask learning model according to claim 1, wherein the specific steps of searching and viewing in step (6) are as follows:
PP1: inputting a primary time period X to be searched by a user through input equipment;
and (3) PP2: inputting the secondary time period X again after the user inputs the primary time period X;
and (3) PP: after the user inputs the primary time period X and the secondary time period X, the computer calls the content required by the user from the XLSX worksheet and displays the content through the display device.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110624171.1A CN113361395B (en) | 2021-06-04 | 2021-06-04 | AI face-changing video detection method based on multitask learning model |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110624171.1A CN113361395B (en) | 2021-06-04 | 2021-06-04 | AI face-changing video detection method based on multitask learning model |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113361395A CN113361395A (en) | 2021-09-07 |
CN113361395B true CN113361395B (en) | 2023-01-17 |
Family
ID=77532213
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110624171.1A Expired - Fee Related CN113361395B (en) | 2021-06-04 | 2021-06-04 | AI face-changing video detection method based on multitask learning model |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113361395B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114339398A (en) * | 2021-12-24 | 2022-04-12 | 天翼视讯传媒有限公司 | Method for real-time special effect processing in large-scale video live broadcast |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106023063A (en) * | 2016-05-09 | 2016-10-12 | 西安北升信息科技有限公司 | Video transplantation face changing method |
WO2019169895A1 (en) * | 2018-03-09 | 2019-09-12 | 华南理工大学 | Fast side-face interference resistant face detection method |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102851208B (en) * | 2012-03-09 | 2015-03-04 | 上海凯度机电科技有限公司 | High-precision cell statistic counting and analyzing device |
CN110929617B (en) * | 2019-11-14 | 2023-05-30 | 绿盟科技集团股份有限公司 | Face-changing synthesized video detection method and device, electronic equipment and storage medium |
CN111343487A (en) * | 2020-03-13 | 2020-06-26 | 深圳市泽云科技有限公司 | Video cloud storage system based on domestic low-consumption processor |
-
2021
- 2021-06-04 CN CN202110624171.1A patent/CN113361395B/en not_active Expired - Fee Related
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106023063A (en) * | 2016-05-09 | 2016-10-12 | 西安北升信息科技有限公司 | Video transplantation face changing method |
WO2019169895A1 (en) * | 2018-03-09 | 2019-09-12 | 华南理工大学 | Fast side-face interference resistant face detection method |
Non-Patent Citations (2)
Title |
---|
Improving the Eddiciency and Robustness of Deepfakes Detection through Precise Geometric Features;Zekun Sun et al;《https://arxiv.org/pdf/2104.04480.pdf》;20210409;第1-10页 * |
基于帧间差异的人脸篡改视频检测方法;张怡暄等;《信息安全学报》;20200315(第02期);第54-77页 * |
Also Published As
Publication number | Publication date |
---|---|
CN113361395A (en) | 2021-09-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US9977952B2 (en) | Organizing images by correlating faces | |
WO2021196389A1 (en) | Facial action unit recognition method and apparatus, electronic device, and storage medium | |
CN106529414A (en) | Method for realizing result authentication through image comparison | |
CN108562589A (en) | A method of magnetic circuit material surface defect is detected | |
CN108830294A (en) | A kind of augmentation method of image data | |
CN104778238B (en) | The analysis method and device of a kind of saliency | |
CN111222433B (en) | Automatic face auditing method, system, equipment and readable storage medium | |
CN110008909A (en) | A kind of real-time audit system of system of real name business based on AI | |
CN104361357B (en) | Photo album categorizing system and sorting technique based on image content analysis | |
CN113361395B (en) | AI face-changing video detection method based on multitask learning model | |
Anichini et al. | The automatic recognition of ceramics from only one photo: The ArchAIDE app | |
CN110058756B (en) | Image sample labeling method and device | |
CN108710893A (en) | A kind of digital image cameras source model sorting technique of feature based fusion | |
Zhang et al. | Improved Fully Convolutional Network for Digital Image Region Forgery Detection. | |
CN110990617B (en) | Picture marking method, device, equipment and storage medium | |
JP2001357067A (en) | Image data retrieving method and computer-readable storage medium | |
CN110363111B (en) | Face living body detection method, device and storage medium based on lens distortion principle | |
CN110807108A (en) | Asian face data automatic collection and cleaning method and system | |
CN108288061A (en) | A method of based on the quick positioning tilt texts in natural scene of MSER | |
CN106341507A (en) | Contact acquiring method, device and user terminal | |
CN114419008A (en) | Image quality evaluation method and system | |
CN109410203A (en) | A kind of picture picture quality detection method based on machine learning | |
CN116308888B (en) | Operation ticket management system based on neural network | |
CN113052234A (en) | Jade classification method based on image features and deep learning technology | |
CN112364824A (en) | Copying detection method of multi-resolution network structure |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20230117 |