CN101354743A - Image base for human face image synthesis - Google Patents
Image base for human face image synthesis Download PDFInfo
- Publication number
- CN101354743A CN101354743A CNA2007100529299A CN200710052929A CN101354743A CN 101354743 A CN101354743 A CN 101354743A CN A2007100529299 A CNA2007100529299 A CN A2007100529299A CN 200710052929 A CN200710052929 A CN 200710052929A CN 101354743 A CN101354743 A CN 101354743A
- Authority
- CN
- China
- Prior art keywords
- image
- template
- face
- people
- library
- 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
Images
Abstract
The invention relates to an image library for the synthesis of face image, which comprises an image library of each face part; a storage structure form of the image library of each face part comprises a coordinate of each feature point of a face part image template and a face part generation image template corresponding to the face part image template; the number of the face part generation image template corresponding to the face part image template is two or more than two; and the storage structure form of the image library of each face part also comprises a corresponding probability value of the face part generation image template corresponding to each face part image template. The image library has two or more than two face part generation image templates corresponding to the face part image template, can automatically generate a plurality of face images and create conditions for improving the similarity degree of an image output by a computer; and the storage structure form comprises the corresponding probability value of the face part generation image template corresponding to each face part image template and creates conditions for the automatic selection of the computer.
Description
Technical field
The present invention relates to the computer image processing technology field, be specifically related to that the automatic generation technique in the automatic generation technique of computer face image field, particularly computer face image uses, be used for the synthetic image library of facial image.
Background technology
The computer image processing technology field, particularly in the automatic generation technique of computer face image field, people's face degradation calculation machine image generating method of appearance, the computing machine generation method automatically of people's face portrait painting etc. after people have invented and have been used to predict the several years.
About people's face degradation calculation machine image generating method:
Appearance can apply to seek lose children on the one hand after being used to predict the several years, improves the performance of face identification system.Aging image under the synthetic on the other hand different weathers can arouse the attention of people to the rational life style of health.The synthetic of the aging image of Realistic Human face also can be used for computer graphics.But people's ageing process itself is a uncertain problem, aging result is not only relevant with people's health, nutrition condition, also be subjected to the influence of external conditions such as living environment,, also have very big challenge so the aging image synthetic technology of people's face has wide application prospect.
The article " reconfiguration technique of the expression of people's face and age conversion and non-complete information " of " electronic letters, vol " the 31st volume 12A phase proposes the aging synthetic method of a kind of people's face based on old and feeble texture and otherness theory.The concrete steps of this method are:
(1) average characteristics of calculating all age group people face;
(2) young facial image is decomposed, obtain corresponding shape and texture vector, be designated as (shp, tex);
(3) calculate new facial image aspect shape and the texture and the difference between the average characteristics, be designated as (Δ shp, Δ tex);
(4) utilize parameter alpha and β to adjust the new facial image and the difference degree of shape facility between the average face and textural characteristics. Δ shp '=α Δ shp, Δ tex '=β Δ tex, 0＜α wherein, β＜1;
(5) after the calculated difference degree changes, the shape vector of new facial image and texture vector (shp ', tex '), shp '=α Δ shp+shp wherein, tex '=β Δ tex+tex. is according to the aging image of the vector forward direction Warp reconstruct that obtains.
The defective of above-mentioned people's face degradation calculation machine image generating method is:
1, the aging result of this synthetic technology depends on the average template that calculates, and all young man's faces are all adopted identical aging mode, therefore, and the poor accuracy of this method; 2, because texture image stack is asked and on average can be produced fuzzyly, the loss detailed information is subjected to resolution limit very big so generate the result, can not synthesize high-resolution image; 3, only consider facial characteristics, do not add consideration, such as hair etc. to influencing other features that the age judges.
Computing machine about people's face portrait painting generates method automatically:
Portrait painting is the facial characteristics that a kind of popular artistic expression, the particularly portrait of people's face can show the individual art uply.But drawing portrait painting is not the born ability of people, has only the artist through long-term train hard just can draw out portrait painting very true to nature.Therefore, how allowing computing machine generate portrait painting automatically according to given facial image obviously is very a difficulty and a challenging problem.
" Chinese journal of computers " the 26th rolled up the 2nd interim article " based on the portrait painting automatic generating calculation of sample learning ", discloses a kind of automatic generation method of portrait painting, and the concrete steps of this method are:
(1), with lineup's face image and corresponding portrait paintings training sample;
(2), the facial image in input detects each characteristic point coordinates;
(3), obtain human face structure, find the point coordinate of facial image correspondence in the training sample by coordinate conversion according to the coordinate setting of each genius loci point of people's face;
(4), according to the corresponding relation of facial image in the training sample and corresponding portrait painting, synthetic and output people face portrait painting with method area sampling of a pixel of a pixel or a zone from training sample of iteration sampling.
Above-mentioned portrait picture automatic generating method defective is:
1, all corresponding identical portrait paintings training sample of face images, therefore, the poor accuracy of this method; 2, owing to the method for sampling of using based on pixel, speed is very slow; 3, when need are exported the portrait painting of another kind of style, this method must take time and effort again according to the training sample generation portrait painting of another kind of style.
Defective in view of above-mentioned people's face degradation calculation machine image generating method, portrait picture automatic generating method existence, also the someone proposes: in the automatic generation technique of computer face image, foundation is used for the synthetic image library of facial image, the shape that each station diagram after will decomposing is then formed as unique point, the shape of forming with each self-corresponding template genius loci point in the portrait painting template base compares respectively, select the template of coupling, it is fast to make the computer face image generate method speed automatically, makes the output image human face similarity degree be improved.
But people imagine the image library of foundation at present, storage organization is simple, generally include only each characteristic point coordinates of each one face's bit image template, the people face position of this people face bit image template correspondence generates image template (as the people face bit image template after aging, the people face position artistic pictures image template of various styles etc.), adopt facial image that the computer face image generation method automatically of above-mentioned image library obtains and imperfect, therefore, at present also there is not a kind of maturation, the computer face image that adopts foundation to be used for facial image composograph storehouse generates method automatically.
Summary of the invention
Technical matters to be solved by this invention is: a kind of synthetic image library of facial image that is used for is provided, and this image library is that the high output facial image of computer face image generation method automatically acquisition similarity creates conditions.
The present invention solves the problems of the technologies described above the technical scheme that is adopted:
Be used for the synthetic image library of facial image, it comprises the image library of each one face position, includes in the storage organization table of the image library of each one face position: the people face position of each characteristic point coordinates of people face bit image template, people face bit image template correspondence generates image template; It is two or more that the people face position of people face bit image template correspondence generates image template, also comprises in the storage organization table of the image library of each one face position: the corresponding probable value of the people face position generation image template of each one face's bit image template correspondence.
In the such scheme, also comprise in the storage organization table of the image library of each one face position: people face position original image template.
In the such scheme, also comprise in the storage organization table of the image library of each one face position: the coding of people face bit image template.
In the such scheme, also comprise in the storage organization table of the image library of each one face position: the age bracket coding.
In the such scheme, each one comprises the image library of face position: hair template image storehouse, eyebrow template image storehouse, eyes template image storehouse, nose template image storehouse, face template image storehouse, face profile template image library.
In the such scheme, each one comprises the image library of face position: wrinkle template image storehouse, senile plaque expelling template image storehouse.
In the such scheme, the resolution of the people face position original image template in hair template image storehouse, the face profile template image library is less than the resolution of the people face position original image template in other image library.
In the such scheme, the resolution of the people face position original image template in wrinkle template image storehouse, the senile plaque expelling template image storehouse is greater than the resolution of the people face position original image template in other image library.
In the such scheme, the people face position of people face bit image template correspondence generates the people face position generation image template that image template is the people face bit image template correspondence of next age bracket.
The advantage of image library of the present invention is:
1, the people face position generation image template of people face bit image template correspondence is two or more, make when computing machine compares, can select the template of a plurality of couplings, can generate a plurality of image facial images automatically, for the similarity that improves computer output image has been created condition.
2, also comprise in the storage organization table of the image library of each one face position: the people face position of each one face's bit image template correspondence generates the corresponding probable value of image template, for condition has been created in the automatic selection of computing machine.Corresponding probable value can also can be obtained by computer learning by artificial setting.
3, also comprise in the storage organization table of the image library of each one face position: people face position original image template.Can make things convenient for computing machine to adopt the image comparison mode to select matching template and create condition.
4, also comprise in the storage organization table of the image library of each one face position: the coding of people face bit image template.Help Computing and storage.
5, also comprise in the storage organization table of the image library of each one face position: the age bracket coding.Behaviour face degradation calculation machine image generates and creates conditions, and helps adding due some characteristics of image of all ages and classes section in the computing machine of people's face portrait painting generates automatically.
6, each one image library of face position comprises: hair template image storehouse, eyebrow template image storehouse, eyes template image storehouse, nose template image storehouse, face template image storehouse, face profile template image library.Help computing machine and select matching template.
7, each one image library of face position comprises: wrinkle template image storehouse, senile plaque expelling template image storehouse.Behaviour face degradation calculation machine image generates and creates conditions, and helps adding due these characteristics of image of all ages and classes section in the computing machine of people's face portrait painting generates automatically.
8, the resolution of the people face position original image template in hair template image storehouse, the face profile template image library is less than the resolution of the people face position original image template in other image library.Under the prerequisite of similarity that does not influence output image and effect, can shorten the working time of computing machine.
9, the people face position of people face bit image template correspondence generates the people face position generation image template that image template is the people face bit image template correspondence of next age bracket.The process matching operation step by step that computing machine can wear out to people's face successively is for the similarity that improves people's face degradation calculation machine output image has been created condition.
Description of drawings
Fig. 1 is the structural representation of the storage organization chained list of image library embodiment 1 of the present invention
Fig. 2 between the image library of each one face position with or figure
Fig. 3 is the eyes template image library storage structural chain list structure synoptic diagram among the image library embodiment 1 of the present invention
Fig. 4 is corresponding relation figure between the eyes template image image storehouse and next age bracket
The behave position view of each unique point of face's bit image template of Fig. 5
Fig. 6 generates the method software flow pattern automatically for the people's face degradation calculation machine image that uses image library embodiment 1 of the present invention
Fig. 7 face ageing process dynamic process figure that behaves
Fig. 8 generates the method output image automatically for the people's face degradation calculation machine image that uses image library embodiment 1 of the present invention
Fig. 9 is the structural representation of the storage organization chained list of image library embodiment 2 of the present invention
Embodiment
The present invention as shown in Figure 1 is used for the synthetic image library embodiment 1 of facial image, and it is the image library of setting up for people's face degradation calculation machine image generating method of appearance after being used to predict the several years.
Embodiment 1 comprises the image library of each one face position, includes in the storage organization table of the image library of each one face position: the people face position of each characteristic point coordinates of people face bit image template, 3 people face bit image template correspondences generates image template, 1 people face position original image template, the coding of people face bit image template, age bracket coding.Also comprise in the storage organization table of the image library at each position of people's face: the people face position of each one face's bit image template correspondence generates the corresponding probable value of image template.It is that the people face position of the people face bit image template correspondence of next age bracket generates image template that the people face position of above-mentioned people face bit image template correspondence generates image template.
Each one comprises the image library of face position: hair template image storehouse, eyebrow template image storehouse, eyes template image storehouse, nose template image storehouse, face template image storehouse, face profile template image library, wrinkle template image storehouse, senile plaque expelling template image storehouse.Relation between the image library of each one face position as shown in Figure 2.
The resolution of the people face position original image template in hair template image storehouse, the face profile template image library is less than the resolution of the people face position original image template in other image library.The resolution of the people face position original image template in wrinkle template image storehouse, the senile plaque expelling template image storehouse is greater than the resolution of the people face position original image template in other image library.
As shown in Figure 3, it is the eyes template image library storage structural chain list structure among the image library embodiment 1.Each characteristic point coordinates of people face bit image template in the pts file, the people face position of 3 people face bit image template correspondences generates image template and corresponding probable value (00000004.jpg.0.55), (0000017.jpg.0.33), (0000009.jpg.0.15), 1 people face position original image template (00008.jpg), the coding (001) of people face bit image template, age bracket coding (40-50).It is that the people face position of the people face bit image template correspondence of next age bracket generates image template that the people face position of above-mentioned people face bit image template correspondence generates image template.
As Fig. 4 is corresponding relation figure between the eyes template image image storehouse and next age bracket, and the arrow thickness is represented the corresponding possibility probability size that exists among the figure.
As shown in Figure 5, according to the position of people's face in the middle of picture, specifically orient the position of eyebrow, eyes, nose, face, face profile.8 unique point coordinate setting of present embodiment 1 usefulness eyebrow, 8 gauge point location eyes, 15 gauge point location noses, 22 gauge point location faces, 25 gauge point location face profiles.That present embodiment 1 adopts is existing face characteristic location technology AAM (Active Appearance Models), can document for reference:
T.F.Cootes，C.J.Taylor，D.Cooper，and?J.Graham，“Active?shape?models-theirtraining?and?application”，Computer?Vision?and?Image?Understanding，61(1)：38-59，1995.
T.F.Cootes，G.J.Edwards?and?C.J.Taylor，“Active?appearance?models”，proceedings?of?ECCV，1998.
People's face degradation calculation machine image as shown in Figure 6 generates the method software flow pattern automatically, and the concrete steps of the automatic generation method of the aging image of people's face are:
(1) input picture:
Provide the user to select the picture of importing and be presented on the interface, the suffix name can be the picture format of .bmp and .jpg etc., and requirement is positive, simple background people face picture, preferably certificate photo.
(2) judge whether people's face:
The picture of judging user's input is people's face, and no, and words require the user to re-enter, and the words that are provide the position of people's face in the middle of picture, cut out.That adopt here is existing human face detection tech (AdaBoost), can document for reference: P.Viola and M.Jones, " Rapid object detection using a boosted cascade of simple features ", CVPR, 2001.
(3) facial image in input detects each characteristic point coordinates, and is each spots localization of people's face.
According to the position of people's face in the middle of picture, specifically orient the position of eyebrow, eyes, nose, face, face profile.
(4) Face Detection:
Detecting skin area exposed in the picture obtains its outline and writes down its position in the middle of picture.Implementation method: at first with picture from the RGB color space conversion to the YCrCb color space; Take out a skin from human face region then and calculate each color component Y, Cr, the mean value of Cb, utilize Cr, the mean value of Cb finds in whole picture and these two pixels that value is approaching, and the pixel that these are approaching becomes white, otherwise become black, we just obtain the bianry image of a black and white like this, skin area be white, other be black; The last profile of white portion that finds in the bianry image of black and white is also with one group of its coordinate position of some record.
(5) background detects:
The background area obtains its outline and writes down its coordinate position in the middle of picture in the detection picture.Implementation method and face detection method are similar, and different is to choose human face region both sides lot in addition to calculate each color component Y, Cr, the mean value of Cb.
(6) hair detects:
Hair zones obtains its outline and writes down its coordinate position in the middle of picture in the detection picture.Implementation method: skin area and background area these two zones are cut apart away in the middle of picture owing to having known, remaining is exactly hair and clothes zone, and what be positioned at the human face region top has been exactly hair zones.
(7), select the image library of coupling according to the position of each unique point of people's face that obtains and each characteristic point coordinates contrast of the people face bit image template among the image library embodiment 1.Be specially:
A. on low resolution, extract the coordinate of facial contour unique point and hair contour feature point, with the unique point coordinate contrast in low resolution face profile template image library and the hair template image storehouse, from aging storehouse, find out only shape of face and hair style sequence, therefrom extract aging mode input picture is carried out burin-in process, obtain the aging image of low resolution;
B. the aging image of low resolution that obtains in the steps A is amplified, on intermediate-resolution, shape facility and textural characteristics according to face, compare with eyebrow template image storehouse, eyes template image storehouse, nose template image storehouse, face template image storehouse in the intermediate-resolution face database, select the aging sequence of only face, be used for the aging result of low resolution is strengthened the face position;
C. the image in random extraction wrinkle template image storehouse, senile plaque expelling template image storehouse on high-rise resolution is used for the aging sequence that step B obtains is strengthened, and replenishes the detailed information of aging sequence.
(8) shape and evolution are carried out to the image in the storehouse of choosing in the characteristic point coordinates position that obtains according to young image; According to the texture informations such as the colour of skin of young image, image in the storehouse is carried out brightness, contrast, the conversion of colourity; Data after the conversion and young man's face are done the fusion editor of image the most at last; Obtain the aging image of high resolving power.
(9) the synthetic aging sequence of people's face of output, as shown in Figure 7, among the figure: (a) classify the young facial image of input as; (b) (c) (d) (e) classifies the aging result of four age brackets in succession as.
Because as time passes, appearance uncertain increasing, therefore, use people's face degradation calculation machine image of image library embodiment 1 of the present invention to generate method software automatically, can obtain one and embody ageing process diversity and probabilistic aging result (as shown in Figure 8).
The present invention is used for the synthetic image library embodiment 2 of facial image, and it is the image library of setting up for the computing machine generation method automatically of people's face portrait painting.
Embodiment 2 comprises the image library of each one face position, and each one comprises the image library of face position: hair template image storehouse, eyebrow template image storehouse, eyes template image storehouse, nose template image storehouse, face template image storehouse, face profile template image library.
The structure of the storage organization chained list of image library embodiment 2 of the present invention includes in the storage organization table of the image library of each one face position as shown in Figure 9: the people face position of each characteristic point coordinates of people face bit image template, three people face bit image template correspondences generates the coding of people's face portrait painting image template and corresponding probable value, people face position original image template, people face bit image template.
Claims (10)
1, is used for the synthetic image library of facial image, it comprises the image library of each one face position, includes in the storage organization table of the image library of each one face position: the people face position of each characteristic point coordinates of people face bit image template, people face bit image template correspondence generates image template; It is characterized in that: the people face position of people face bit image template correspondence generates image template and is two or more, also comprises in the storage organization table of the image library at each position of people's face: the people face position of each one face's bit image template correspondence generates the corresponding probable value of image template.
2, image library as claimed in claim 1 is characterized in that: also comprise in the storage organization table of the image library of each one face position: people face position original image template.
3, image library as claimed in claim 1 is characterized in that: also comprise in the storage organization table of the image library of each one face position: the coding of people face bit image template.
4, image library as claimed in claim 1 is characterized in that: also comprise in the storage organization table of the image library of each one face position: the age bracket coding.
5, as claim 1 or 3 or 4 described image libraries, it is characterized in that: each one comprises the image library of face position: hair template image storehouse, eyebrow template image storehouse, eyes template image storehouse, nose template image storehouse, face template image storehouse, face profile template image library.
6, image library as claimed in claim 2 is characterized in that: each one comprises the image library of face position: hair template image storehouse, eyebrow template image storehouse, eyes template image storehouse, nose template image storehouse, face template image storehouse, face profile template image library.
7, image library as claimed in claim 6 is characterized in that: each one comprises the image library of face position: wrinkle template image storehouse, senile plaque expelling template image storehouse.
8, image library as claimed in claim 7 is characterized in that: the resolution of the people face position original image template in hair template image storehouse, the face profile template image library is less than the resolution of the people face position original image template in other image library.
9, image library as claimed in claim 7 is characterized in that: the resolution of the people face position original image template in wrinkle template image storehouse, the senile plaque expelling template image storehouse is greater than the resolution of the people face position original image template in other image library.
10, image library as claimed in claim 4 is characterized in that: the people face position of people face bit image template correspondence generates the people face position generation image template that image template is the people face bit image template correspondence of next age bracket.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2007100529299A CN101354743A (en) | 2007-08-09 | 2007-08-09 | Image base for human face image synthesis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2007100529299A CN101354743A (en) | 2007-08-09 | 2007-08-09 | Image base for human face image synthesis |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101354743A true CN101354743A (en) | 2009-01-28 |
Family
ID=40307549
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2007100529299A Pending CN101354743A (en) | 2007-08-09 | 2007-08-09 | Image base for human face image synthesis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101354743A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103955708A (en) * | 2014-05-13 | 2014-07-30 | 重庆大学 | Face photo library fast-reduction method for face synthesis portrait recognition |
| CN105096353A (en) * | 2014-05-05 | 2015-11-25 | 腾讯科技（深圳）有限公司 | Image processing method and device |
| CN105654420A (en) * | 2015-12-21 | 2016-06-08 | 小米科技有限责任公司 | Face image processing method and device |
| CN108463823A (en) * | 2016-11-24 | 2018-08-28 | 华为技术有限公司 | A kind of method for reconstructing, device and the terminal of user's Hair model |
| CN109034122A (en) * | 2018-08-29 | 2018-12-18 | 吴伟锋 | Voice-control nose analysis platform |
| CN109359626A (en) * | 2018-11-21 | 2019-02-19 | 合肥金诺数码科技股份有限公司 | A kind of Image Acquisition complexion prediction meanss and its prediction technique |
| CN110222588A (en) * | 2019-05-15 | 2019-09-10 | 合肥进毅智能技术有限公司 | A kind of human face sketch image aging synthetic method, device and storage medium |
| CN111209050A (en) * | 2020-01-10 | 2020-05-29 | 北京百度网讯科技有限公司 | Method and device for switching working mode of electronic equipment |
| CN112037135A (en) * | 2020-09-11 | 2020-12-04 | 上海瞳观智能科技有限公司 | Method for selecting image key main body to be amplified and displayed |
-
2007
- 2007-08-09 CN CNA2007100529299A patent/CN101354743A/en active Pending
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105096353A (en) * | 2014-05-05 | 2015-11-25 | 腾讯科技（深圳）有限公司 | Image processing method and device |
| CN103955708A (en) * | 2014-05-13 | 2014-07-30 | 重庆大学 | Face photo library fast-reduction method for face synthesis portrait recognition |
| CN103955708B (en) * | 2014-05-13 | 2017-01-25 | 重庆大学 | Face photo library fast-reduction method for face synthesis portrait recognition |
| CN105654420A (en) * | 2015-12-21 | 2016-06-08 | 小米科技有限责任公司 | Face image processing method and device |
| CN108463823A (en) * | 2016-11-24 | 2018-08-28 | 华为技术有限公司 | A kind of method for reconstructing, device and the terminal of user's Hair model |
| CN109034122A (en) * | 2018-08-29 | 2018-12-18 | 吴伟锋 | Voice-control nose analysis platform |
| CN109359626A (en) * | 2018-11-21 | 2019-02-19 | 合肥金诺数码科技股份有限公司 | A kind of Image Acquisition complexion prediction meanss and its prediction technique |
| CN110222588A (en) * | 2019-05-15 | 2019-09-10 | 合肥进毅智能技术有限公司 | A kind of human face sketch image aging synthetic method, device and storage medium |
| CN111209050A (en) * | 2020-01-10 | 2020-05-29 | 北京百度网讯科技有限公司 | Method and device for switching working mode of electronic equipment |
| CN112037135A (en) * | 2020-09-11 | 2020-12-04 | 上海瞳观智能科技有限公司 | Method for selecting image key main body to be amplified and displayed |
| CN112037135B (en) * | 2020-09-11 | 2023-06-09 | 上海瞳观智能科技有限公司 | Method for magnifying and displaying selected image key main body |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10535163B2 (en) | Avatar digitization from a single image for real-time rendering | |
| CN101354743A (en) | Image base for human face image synthesis | |
| Lin | Face detection in complicated backgrounds and different illumination conditions by using YCbCr color space and neural network | |
| CN101477696B (en) | Human character cartoon image generating method and apparatus | |
| CN107944428B (en) | Indoor scene semantic annotation method based on super-pixel set | |
| Zhang et al. | Style transfer via image component analysis | |
| WO2019050808A1 (en) | Avatar digitization from a single image for real-time rendering | |
| JP2010507854A (en) | Method and apparatus for virtual simulation of video image sequence | |
| CN107945244A (en) | A kind of simple picture generation method based on human face photo | |
| CN105718885B (en) | A kind of Facial features tracking method | |
| Suo et al. | High-resolution face fusion for gender conversion | |
| KR20230097157A (en) | Method and system for personalized 3D head model transformation | |
| JP2017102622A (en) | Image processing device, image processing method and program | |
| CN110533579A (en) | Based on the video style conversion method from coding structure and gradient order-preserving | |
| CN110910479B (en) | Video processing method, device, electronic equipment and readable storage medium | |
| Sýkora et al. | Sketching Cartoons by Example. | |
| Zhang et al. | Imageadmixture: Putting together dissimilar objects from groups | |
| Luo et al. | Facial metamorphosis using geometrical methods for biometric applications | |
| US11361189B2 (en) | Image generation method and computing device | |
| Lin et al. | Video stylization: painterly rendering and optimization with content extraction | |
| Kawai et al. | Data-driven speech animation synthesis focusing on realistic inside of the mouth | |
| Milevskiy et al. | A fast algorithm for korean text extraction and segmentation from subway signboard images utilizing smartphone sensors | |
| Mavzuna | MODELING OF TEXT RECOGNITION IN IMAGES | |
| Pourjam et al. | Segmentation of Human Instances Using Grab-cut and Active Shape Model Feedback. | |
| Gómez-Mendoza | Automatic lip-contour extraction and mouth-structure segmentation in images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Open date: 20090128 |