CN116385584A - Poster generation method, device and system and computer readable storage medium - Google Patents
Poster generation method, device and system and computer readable storage medium Download PDFInfo
- Publication number
- CN116385584A CN116385584A CN202310349603.1A CN202310349603A CN116385584A CN 116385584 A CN116385584 A CN 116385584A CN 202310349603 A CN202310349603 A CN 202310349603A CN 116385584 A CN116385584 A CN 116385584A
- Authority
- CN
- China
- Prior art keywords
- poster
- template
- text
- prompt
- demand
- 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
- 238000000034 method Methods 0.000 title claims abstract description 64
- 238000013461 design Methods 0.000 claims abstract description 65
- 238000012545 processing Methods 0.000 claims abstract description 27
- 239000013598 vector Substances 0.000 claims description 27
- 238000013500 data storage Methods 0.000 claims description 12
- 230000011218 segmentation Effects 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 6
- 239000003607 modifier Substances 0.000 claims description 5
- 238000004891 communication Methods 0.000 claims description 2
- 238000009877 rendering Methods 0.000 claims description 2
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 19
- 238000004519 manufacturing process Methods 0.000 description 15
- 238000004590 computer program Methods 0.000 description 11
- 238000004422 calculation algorithm Methods 0.000 description 8
- 238000013527 convolutional neural network Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 238000009792 diffusion process Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 230000006872 improvement Effects 0.000 description 3
- 238000012015 optical character recognition Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 241000282326 Felis catus Species 0.000 description 2
- 238000003491 array Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 239000003795 chemical substances by application Substances 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 230000004992 fission Effects 0.000 description 2
- 238000003709 image segmentation Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 241001050985 Disco Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000001143 conditioned effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000011022 operating instruction Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/001—Texturing; Colouring; Generation of texture or colour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/12—Use of codes for handling textual entities
- G06F40/126—Character encoding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/284—Lexical analysis, e.g. tokenisation or collocates
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/20—Drawing from basic elements, e.g. lines or circles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/60—Editing figures and text; Combining figures or text
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/23—Processing of content or additional data; Elementary server operations; Server middleware
- H04N21/235—Processing of additional data, e.g. scrambling of additional data or processing content descriptors
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/80—Generation or processing of content or additional data by content creator independently of the distribution process; Content per se
- H04N21/83—Generation or processing of protective or descriptive data associated with content; Content structuring
- H04N21/84—Generation or processing of descriptive data, e.g. content descriptors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2200/00—Indexing scheme for image data processing or generation, in general
- G06T2200/24—Indexing scheme for image data processing or generation, in general involving graphical user interfaces [GUIs]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2200/00—Indexing scheme for image data processing or generation, in general
- G06T2200/32—Indexing scheme for image data processing or generation, in general involving image mosaicing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application relates to the technical field of artificial intelligence and discloses a poster generation method, device and system and a computer readable storage medium, wherein the method comprises the following steps: acquiring a poster template, extracting image information of the poster template, and generating a template prompt text according to the image information; acquiring at least one item of demand information according to a preset rule, and generating a demand prompt text according to the demand information; and inputting the template prompt text and the demand prompt text into the intelligent poster generation model for processing to obtain a poster picture. According to the method and the device, the template prompt text can be generated by acquiring the poster template, the personalized demand prompt text is generated according to the demand of the user, the template prompt text and the demand prompt text are input into the intelligent poster generation model together, and the poster generation flow is engineized, so that the poster threshold is greatly reduced; meanwhile, the generation of the poster picture can be completed with low cost, high efficiency and high quality without depending on excessive design resources.
Description
Technical Field
The present disclosure relates to the field of artificial intelligence technologies, and in particular, to a method, an apparatus, a system, and a computer readable storage medium for generating a poster.
Background
As a common poster form, the poster is widely used in advertising and other scenes, and the poster with various forms plays an indispensable important role in on-line marketing fission activities. In designing a poster, designers often pay considerable time, such as modifying the content of a document, designing the format of the poster, expanding multiple dimensions for a display, etc., which can consume a great deal of time and labor cost, and meanwhile, the poster needs to present the effect of thousands of people and thousands of faces under the high-flow background that the accurate marketing is a trend. This places very high demands on the production efficiency of the poster.
In the prior art, the poster can be generated by a poster generating tool, and the prior art relies on a large amount of design resources, so that the production efficiency of the poster still has a further improvement space under the condition that the design resources are limited or the requirements of the poster are not met. Thus, there is a need for a method that can quickly generate high quality posters.
Disclosure of Invention
In view of the above, embodiments of the present application provide a method, an apparatus, a system, and a computer readable storage medium for generating a poster, which aim to solve the problem of low production efficiency of the poster.
In a first aspect, an embodiment of the present application provides a method for generating a poster, where the method is applied to a server side of a poster generating system; the method comprises the following steps:
acquiring a poster template, extracting image information of the poster template, and generating a template prompt text according to the image information;
acquiring at least one item of demand information according to a preset rule, and generating a demand prompt text according to the demand information, wherein the preset rule is used for describing a plurality of personalized parameters required for generating a poster;
and inputting the template prompt text and the demand prompt text into a poster intelligent generation model for processing to obtain a poster picture.
In a second aspect, an embodiment of the present application further provides a device for generating a poster, where the device is used for a server side of a poster generating system; the device comprises:
the first text generation module is used for acquiring a poster template, extracting image information of the poster template and generating a template prompt text according to the image information;
the second text generation module is used for acquiring at least one item of demand information according to a preset rule and generating a demand prompt text according to the demand information, wherein the preset rule is used for describing a plurality of personalized parameters required for generating the poster;
And the generating module is used for inputting the template prompt text and the demand prompt text into a poster intelligent generating model for processing to obtain a poster picture.
In a third aspect, an embodiment of the present application further provides a poster generation system, where the system includes a client and a server that are connected in a communication manner, where the server is deployed with the poster generation device described above;
the client is used for displaying a front-end interface, receiving the demand information through the front-end interface, sending the demand information to the server, receiving the poster image sent by the server and rendering.
In a fourth aspect, embodiments of the present application further provide an electronic device, including: a processor; and a memory arranged to store computer executable instructions that, when executed, cause the processor to perform the steps of the method of generating a poster as described in any of the above.
In a fifth aspect, embodiments of the present application further provide a computer-readable storage medium storing one or more programs that, when executed by an electronic device including a plurality of application programs, cause the electronic device to perform the steps of any of the above-described methods of generating a poster.
The above-mentioned at least one technical scheme that this application embodiment adopted can reach following beneficial effect:
according to the generation method of the poster, the image information of the poster template is extracted, the template prompt text is generated according to the image information, one or more pieces of requirement information are obtained according to preset rules, the requirement prompt text is generated according to the requirement information, and finally, the template prompt text and the requirement prompt text are input into the intelligent generation model of the poster for processing, so that the poster picture is obtained. It can be seen that the template prompt text can be generated by acquiring the poster template, meanwhile, the personalized demand prompt text can be generated according to the demand of the user, the template prompt text and the demand prompt text are input into the intelligent poster generation model together, the poster generation flow is engineized, the poster threshold is greatly reduced, and non-design personnel such as products, operation and the like have the capability of manufacturing the poster; meanwhile, the generation of the poster picture can be completed with low cost, high efficiency and high quality without depending on excessive design resources.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute an undue limitation to the application. In the drawings:
Fig. 1 shows a schematic structural diagram of a poster generating system according to one embodiment provided herein;
fig. 2 shows a flow diagram of a method for generating a poster according to one embodiment provided herein;
FIG. 3 illustrates a schematic diagram of the structure of a poster template according to one embodiment provided herein;
FIG. 4 illustrates a schematic diagram of a poster intelligent generation model, according to one embodiment provided herein;
fig. 5 shows a flow diagram of a method for generating a poster according to another embodiment provided herein;
fig. 6 shows a schematic structural diagram of a poster generating apparatus according to an embodiment provided herein;
fig. 7 shows a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
For the purposes, technical solutions and advantages of the present application, the technical solutions of the present application will be clearly and completely described below with reference to specific embodiments of the present application and corresponding drawings. It will be apparent that the described embodiments are only some, but not all, of the embodiments of the present application. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.
It should be noted that the terms "first," "second," and the like in the description and claims of the present application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that such uses may be interchanged where appropriate such that embodiments of the present application described herein may be implemented in sequences other than those illustrated or described herein. Furthermore, the terms "include" and variations thereof are to be interpreted as open-ended terms that mean "include, but are not limited to.
The following describes in detail the technical solutions provided by the embodiments of the present application with reference to the accompanying drawings.
As a common poster form, the poster is widely used in advertising and other scenes, and the poster with various forms plays an indispensable important role in on-line marketing fission activities. In designing a poster, designers often pay considerable time, such as modifying the content of a document, designing the format of the poster, expanding multiple dimensions for a display, etc., which can consume a great deal of time and labor cost, and meanwhile, the poster needs to present the effect of thousands of people and thousands of faces under the high-flow background that the accurate marketing is a trend. This places very high demands on the production efficiency of the poster. In the prior art, the poster can be generated by a poster generating tool, and the prior art relies on a large amount of design resources, so that the production efficiency of the poster still has a further improvement space under the condition that the design resources are limited or the requirements of the poster are not met.
Based on the method, the template prompt text can be generated by acquiring the poster template, meanwhile, the personalized demand prompt text can be generated according to the demand of a user, the template prompt text and the demand prompt text are input into the intelligent poster generation model together, the poster generation flow is engineized, the poster threshold is greatly reduced, and non-designer personnel such as products, operation and the like have the capability of manufacturing the poster; meanwhile, the generation of the poster picture can be completed with low cost, high efficiency and high quality without depending on excessive design resources.
Fig. 1 shows a schematic structural diagram of a poster generating system according to an embodiment provided in the present application, and the generating method of the poster provided in the embodiment of the present application may be implemented in a system as shown in fig. 1, and as can be seen from fig. 1, the poster generating system 100 includes a client 101 and a server 102 that are communicatively connected, where the server 102 is deployed with a generating device 600 (fig. 6) of a poster. The generation of the poster of the present application is specifically applied to the server 102 of the poster generation system 100. The client 101 is configured to display a front-end interface, receive the demand information through the front-end interface, send the demand information to the server 102, and receive and render the poster image sent by the server 102.
It should be noted that, the present application is not limited to the poster generation system shown in fig. 1, and any system, device, framework, etc. capable of implementing the business logic of the present application may be used, and fig. 1 is only an exemplary illustration.
Fig. 2 is a flow chart illustrating a method for generating a poster according to an embodiment provided in the present application, and as can be seen from fig. 2, the present application at least includes steps S201 to S203:
step S201: and acquiring a poster template, extracting image information of the poster template, and generating a template prompt text according to the image information.
First, a poster template is acquired. Specifically, in some embodiments of the present application, in the above method, the obtaining a poster template includes: obtaining a graphic design file, and taking the graphic design file as the poster template; or, obtaining a plurality of configuration elements of the poster template, typesetting the plurality of configuration elements to generate the poster template, wherein the configuration elements comprise basic elements and business elements, the basic elements comprise at least one of lines, geometric figures, pictures and texts, and the business elements comprise at least one of two-dimensional codes, applet codes and head portraits.
In this embodiment, the obtained graphic design file may be directly passed through as a poster template. For example, the graphic design files include, but are not limited to, PSD files edited by image processing software Photoshop, AI files edited by vector graphic authoring software Adobe Illustrator, CDR files edited by flat design software coredraw. When in implementation, a user can transmit local graphic design files to a server in real time through a network, and the size and the number of the files are not limited. According to the embodiment, the manufactured poster template is transmitted to the server side in a local uploading mode, so that the cost of manufacturing the poster template in the early stage is reduced, the subsequent server side can generate a final poster picture based on the poster template, and the production efficiency of the poster template is improved.
In this embodiment, the poster template may also be generated by obtaining a plurality of configuration elements of the poster template, and typesetting the plurality of configuration elements. The configuration element comprises a basic element and a business element, wherein the basic element comprises at least one of lines, geometric figures, pictures and texts, and the business element comprises at least one of two-dimensional codes, small Cheng Xuma and head images. For example, fig. 3 shows a schematic structural diagram of a poster template according to an embodiment provided in the present application, from which it can be seen that the following configuration elements can be obtained: typesetting the 5 configuration elements by using the text 1, the text 2, the text 3, the geometric figure and the two-dimensional code can generate the poster template shown in fig. 3. According to the embodiment, the user can typeset the configuration elements by acquiring the configuration elements, so that the poster template is generated, the user can generate the poster template by the method without mastering professional knowledge, the poster template is simple and quick, and the poster template can be designed according to actual needs so as to generate a final poster picture based on the poster template; or directly taking the poster template meeting the actual requirements as a final poster picture, thereby improving the production efficiency of the poster template.
After the poster template is obtained, the image information of the poster template can be extracted, and a template prompt text is generated according to the image information. Specifically, in some embodiments, the poster template is generated by typesetting configuration elements, and then the Region-based Mask convolutional neural network model (Mask Region-based Convolutional Neural Networks, mask R-CNN) can be used to extract image information of the poster template, and the poster template is subjected to example segmentation to obtain a plurality of labels, and the labels are used as template prompt texts.
From the above embodiments, it can be seen that, by directly uploading the graphic design file or directly typesetting the configuration elements, the poster template can be obtained, so that the manufacturing of the poster threshold is greatly reduced, and the poster picture can be generated on the basis of the poster template.
Step S202: at least one item of demand information is obtained according to a preset rule, and a demand prompt text is generated according to the demand information, wherein the preset rule is used for describing a plurality of personalized parameters required for generating the poster.
At least one item of demand information is obtained according to a preset rule, and a demand prompt text is generated according to the demand information, wherein the preset rule is used for describing a plurality of personalized parameters required for generating the poster. In some embodiments, a sentence may be obtained: a cat riding can generate a demand prompt text according to the demand information: car, cat.
In other embodiments of the present application, in the foregoing method, in the preset rule, the input form of the plurality of personalized parameters is: the plurality of personalization parameters = [ picture subject ] (or [ adjective ] [ subject ]), [ detail setting ], [ modifier or artist ]; the obtaining at least one piece of demand information according to a preset rule includes: displaying a front-end interface on a client, wherein the front-end interface is generated according to the preset rule, the front-end interface comprises a plurality of personalized parameter controls, and the personalized parameter controls are associated with a pre-built recommended label library; and responding to the input operation of the user on the front-end interface, and acquiring the requirement information.
In this embodiment, a front-end interface may be displayed on the client, where the front-end interface is generated according to a preset rule, and a user may input a plurality of personalized parameters on the front-end interface, where the input form of the plurality of personalized parameters in the preset rule is: the plurality of personalization parameters = [ picture subject ] (or [ adjective ] [ subject ]), [ detail setting ], [ modifier or artist ]. In practice, the personalization parameters further include: resolution size, picture detail parameters, etc., to which the present application is not limited.
And responding to the input operation of the user on the front-end interface, acquiring the requirement information, and generating a requirement prompt text according to the requirement information. In some embodiments, for example, through a picture body parameter control, a user may input a personalized parameter "person" that may be used directly as the demand prompt text. For another example, through the detail setting parameter control, a user can input a picture, the picture can be used as requirement information, feature matching is performed on the picture, a text 1 and a text 2 are obtained, and the text 1 and the text 2 can be used as requirement prompt texts. For another example, a user may input a text through a modifier or an artist parameter control, and may use the text as requirement information, and because the personalized parameter control in this embodiment may be associated with a pre-built recommended tag library, the server may analyze the requirement information, select a recommended tag related to the text from the pre-built recommended tag library, and recommend the recommended tag to the user, and the user may select a plurality of tags as required by himself or herself, as requirement prompt text.
In other embodiments, pre-options may be set in the personalized parameter control for selection by the user. For example, for the picture body parameter control described above, the following 5 options may be set: a person, cat, building, football, flower, if the user wishes to create a poster template with a single building, then the building's option can be selected.
In some embodiments of the present application, in the above method, the recommended tag library is constructed according to the following method: obtaining a plurality of text samples, and carrying out sentence segmentation on each text sample to obtain a plurality of sub-sentences; performing word segmentation and part-of-speech tagging on each sub-sentence, and reserving a plurality of candidate words with appointed parts-of-speech; constructing a candidate keyword graph according to the reserved multiple candidate words so as to determine the weight of each candidate word; and according to the weight, descending order is carried out on the candidate words, and a plurality of candidate words which are ranked in front are intercepted to be used as labels, so that the recommended label library is formed.
The text sample described in this embodiment may be text of a text input by a user of another poster generating system. In this embodiment, a text sample 1, a text sample 2, and a text sample 3 may be acquired. Sentence segmentation is carried out on the text sample 1 to obtain a sub sentence 11 and a sub sentence 12; sentence segmentation is carried out on the text sample 2 to obtain a sub sentence 21 and a sub sentence 22; sentence segmentation is performed on the text sample 3 to obtain sub-sentences 31 and sub-sentences 32.
Taking the sub-sentence 11 as an example, performing word segmentation and part-of-speech labeling on the sub-sentence 11 to obtain a candidate word 1, a candidate word 2, a candidate word 3, a candidate word 4 and a candidate word 5, wherein the part-of-speech of the candidate word 1 is a measuring word, the parts-of-speech of the candidate word 2 and the candidate word 4 is a noun, and the parts-of-speech of the candidate word 3 and the candidate word 5 is an adjective, and if the part-of-speech is a noun and an adjective, the candidate word 2, the candidate word 3, the candidate word 4 and the candidate word 5 can be reserved. According to the reserved candidate words, a candidate keyword graph G= (V, E) is constructed, wherein V is a node set consisting of a candidate word 2, a candidate word 3, a candidate word 4 and a candidate word 5, and the weight of each candidate word is determined according to a similarity calculation formula of a text ranking algorithm TextRank algorithm, for example, the weight corresponding to the candidate word 2 is determined to be 0.1, the weight corresponding to the candidate word 3 is determined to be 0.5, the weight corresponding to the candidate word 4 is determined to be 0.1, and the weight corresponding to the candidate word 5 is determined to be 0.3. And (3) descending order arrangement is carried out on each candidate word according to the weight, so that an ordering result is obtained: candidate 3, candidate 5, candidate 2, candidate 4, the top 2 candidates may be ranked: candidate word 3 and candidate word 5 are used as tag 1 and tag 2.
The above steps are performed for the sub-sentence 12, the sub-sentence 21, the sub-sentence 22, the sub-sentence 31, and the sub-sentence 32, respectively, to obtain the label 3, the label 4, the label 5, the label 6, the label 7, the label 8, the label 9, and the label 10, and a recommended label library is formed.
According to the embodiment, important keywords are extracted from the text of other users of the poster generation system, so that a recommendation tag library can be precipitated, and when the current user uses the poster generation system, recommendation tags in the recommendation tag library can be recommended to the current user for reference by the current user, and user experience can be improved; the demand prompt text is rapidly determined, so that the production efficiency of poster pictures can be improved.
Step S203: and inputting the template prompt text and the demand prompt text into a poster intelligent generation model for processing to obtain a poster picture.
And finally, the template prompt text and the demand prompt text can be input into the intelligent poster generation model for processing, so as to obtain a poster picture. Specifically, in some embodiments, the template prompt text and the demand prompt text may be spliced to obtain a spliced prompt text, and the spliced prompt text is input into a Diffusion model Disco Diffusion model to be processed to obtain a plurality of poster pictures, where the number of generated poster pictures may be set according to actual needs, and the disclosure is not limited. According to the embodiment, the template prompt text and the demand prompt text are input into the intelligent poster generation model for processing, so that a plurality of poster pictures can be obtained, and when business data change, a user can quickly generate the poster pictures by modifying the template prompt text and the demand prompt text, and compared with a poster generation tool in the prior art, the time cost and the labor cost are reduced; the poster pictures can be generated in batches, and the production efficiency of the poster is improved.
As can be seen from the method shown in fig. 1, in the method for generating a poster provided by the present application, by acquiring a poster template, extracting image information of the poster template, generating a template prompt text according to the image information, acquiring one or more items of requirement information according to a preset rule, generating a requirement prompt text according to the requirement information, and finally, inputting the template prompt text and the requirement prompt text into a smart generation model of the poster for processing, thereby obtaining a poster picture. It can be seen that the template prompt text can be generated by acquiring the poster template, meanwhile, the personalized demand prompt text can be generated according to the demand of the user, the template prompt text and the demand prompt text are input into the intelligent poster generation model together, the poster generation flow is engineized, the poster threshold is greatly reduced, and non-design personnel such as products, operation and the like have the capability of manufacturing the poster; meanwhile, the generation of the poster picture can be completed with low cost, high efficiency and high quality without depending on excessive design resources.
In some embodiments of the present application, in the above method, the poster template is a graphic design file; the extracting the image information of the poster template, generating a template prompt text according to the image information, comprises the following steps: extracting target prompt information from the poster template according to the file type of the graphic design file; if the graphic design file is a first format file, extracting at least one of a file header, detailed color mode information, image source data, image layer and mask data and image information of the graphic design file according to a data storage format of the graphic design file; if the graphic design file is a file in a second format, extracting multi-level information from the graphic design file according to a data storage format of the graphic design file, wherein coarse-grained information in the multi-level information comprises at least one of beginning description, indirect objects, a cross index table and file tail; and executing at least one of character recognition, layer picture source feature matching and global picture feature matching on the target prompt information of each item to generate the template prompt text, wherein the template prompt text is one or more of global prompt text, layer picture source prompt text and literal prompt text.
In this embodiment, the target prompt information may be extracted from the poster template according to the file type of the graphic design file. The following description will take a graphic design file with a file type as a PSD format and an AI format as an example.
If the graphic design file is a PSD file, in some embodiments, a header of the graphic design file may be extracted according to a data storage format of the graphic design file, where the header includes basic information of the PSD file, such as a version number, a transparent channel number, a width, a height, a depth, a color mode, and the like. In other embodiments, the header of the graphic design file and detailed color mode information may be extracted according to a data storage format of the graphic design file, wherein the detailed color mode information includes color data Colordata, and a length of the color data. In still other embodiments, the header, detailed color mode information, source data, layer and mask data, and picture information of the graphic design file may be extracted according to a data storage format of the graphic design file, where the source data includes a picture resource and a length of a picture resource portion; the layer and mask data comprises miscellaneous information length, layer information and global layer mask information; the picture information includes picture data and a compression format of the picture.
If the graphic design file is a file in an AI format, the graphic design file can be subjected to multi-level information extraction according to the data storage format of the graphic design file to obtain one or more of a beginning description, an indirect object, a cross index table and a file tail. Wherein the cross index table holds offset addresses of all obj's in the file, all data information is held in obj's, and each obj object is similar to the structure of table 1 below.
TABLE 1
| Comments |
| Objm |
| … |
| Objn |
| xref |
| Trailer(rootIndex) |
| startxref |
| ％％EOF |
Wherein Comments are the beginning of the document; objm, …, objn are indirect objects, the object order being crossed; xref lists the file offset addresses of all indirect objects by row; trailer (rootIndex), including file information, including root directory object index, info directory object index, total number of indirect objects; startxref contains the indirect object cross index table file offset address.
Table 2 below shows the content format of the graphic design file in AI format:
TABLE 2
The data are stored in the Obj corresponding to the file AIPrivatedata in blocks according to the sequence, wherein the length of the Obj data is 65536 bytes at most, and the data can be divided into: compressed (text file), uncompressed (text file). The above Objects part exists in the layer information, path data, and image data. When the multi-level information extraction is carried out on the graphic design file, the beginning of the layer data can be searched, and then the specific path data is read and analyzed according to the rows until the data area is finished and the Page tracker is started.
After the target prompt information is extracted, at least one of character recognition, layer image source feature matching and global image feature matching can be performed on each target prompt information, and a template prompt text is generated. Specifically, in some embodiments, optical character recognition (Optical Character Recognition, OCR) may be performed on a layer diagram source in the target hint information from a PSD format file or an AI format file to obtain literal hint text. In other embodiments, feature matching may be performed on the layer source in the target hint information from the PSD format file, for example, feature extraction may be performed on the layer source using a Mask R-CNN model to obtain a plurality of texts, where the plurality of texts are used as layer source hint texts. In still other embodiments, a Mask R-CNN model may be used to perform feature matching on a global picture in the target prompt message from the PSD format file or the AI format file, so as to obtain a plurality of texts, and the plurality of texts are used as global prompt texts. In still other embodiments, the layer name may also be used directly as a literal hint text.
In some embodiments of the present application, before the step of inputting the template prompt text and the requirement prompt text into the intelligent poster generation model for processing, the method further includes: performing de-duplication processing on the template prompt text and the required prompt text to obtain a de-duplicated prompt text; according to the degree of association with the demand information, the de-duplicated prompt texts are ordered in a descending order to obtain ordered prompt texts; and according to the sorting order, assigning a weight to each text in the sorted prompt texts so as to preprocess the template prompt texts and the demand prompt texts.
In this embodiment, the template prompt text and the demand prompt text may be preprocessed first, and then the preprocessed text is input into the intelligent poster generation model for processing. Specifically, if the template prompt text comprises lawns, equipment, mountains and sunlight, the demand prompt text comprises equipment, leases, factories, automobile trade and automobile articles, the template prompt text and the demand prompt text can be subjected to duplication removal treatment to obtain a duplication-removed prompt text comprising lawns, equipment, mountains, sunlight, leases, factories, automobile trade and automobile articles; then, if the demand information is set as the automobile trade, the de-duplicated prompt texts can be ordered in a descending order according to the degree of association with the automobile trade, so that the ordered prompt texts are obtained: trade, lease, automotive, factory, equipment, lawn, mountain, sun; finally, according to the ranking order, each text in the ranked prompt texts is assigned with a weight, for example, the weight of a 'car trade' text is 0.3, the weight of a 'leasing' text is 0.15, the weight of a 'car articles' text is 0.13, the weight of a 'factory' text is 0.12, the weight of a 'device' text is 0.1, the weight of a 'lawn' text is 0.08, the weight of a 'mountain' text is 0.07, the weight of a 'sunshine' text is 0.05, each text is multiplied with the corresponding weight, so that the template prompt text and the demand prompt text are preprocessed, the preprocessed text is obtained, and the preprocessed text can be input into a poster intelligent generation model for processing.
In some embodiments of the present application, in the above method, the poster intelligent generation model is constructed by a Stable diffration model, and the poster intelligent generation model includes a text encoding layer, an image information creating layer and an image decoding layer connected in sequence; inputting the template prompt text and the demand prompt text into a poster intelligent generation model for processing to obtain a poster picture, wherein the method comprises the following steps of: coding the template prompt text and the demand prompt text based on the text coding layer to obtain text vectors; denoising random noise data according to the text vector based on the image information creation layer to obtain an image vector; and decoding the image vector based on the image decoding layer to obtain the poster picture.
The Stable diffration model is a text-to-image model, and is mainly used for denoising random gaussian noise according to the description of the text so as to generate an image.
Fig. 4 shows a schematic structural diagram of a smart poster generation model according to one embodiment provided in the present application, and as can be seen from fig. 4, the smart poster generation model includes a text encoding layer 401, an image information creating layer 402, and an image decoding layer 403, which are sequentially connected. The present embodiment is exemplarily described below with reference to fig. 4.
First, a template prompt text and a demand prompt text may be spliced to obtain a spliced prompt text, the spliced prompt text is encoded by using a text encoding layer 401 to obtain a text vector, for example, a text encoder of a contrast language-Image Pre-training model (Contrastive Language-Image Pre-training, CLIP) may be used to encode the spliced prompt text to obtain the text vector, where the dimension of the text vector may be 77×768, which is not limited in this application.
Then, generating random noise data, wherein the random noise data can be noise data formed by initialized multidimensional arrays; using the image information creation layer 402, conditioned on text vectors, wherein the image information creation layer may include a U-Net model and a scheduling algorithm, wherein the scheduling algorithm may be, but is not limited to, denoising random noise data to obtain denoised data 1, for example, a denoising diffusion implicit model (Denoising diffusion implicit models, DDIM), a kernel minimum mean square error algorithm (Kernel Least Mean Square, KLMS), etc.; denoising the denoising data 1 by using a U-Net model and a scheduling algorithm under the condition of the text vector to obtain denoising data 2; denoising the denoising data 2 by using a convolution network (Convolutional Networks for Biomedical Image Segmentation, U-Net) model for biomedical image segmentation, a DDIM algorithm and the like under the condition of the text vector to obtain denoising data 3; and by analogy, denoising the denoising data N-1 by using a U-Net model and a scheduling algorithm under the condition of the text vector to obtain an image vector, wherein N is a positive integer, and the value of N can be set according to actual needs, for example, the value of N can be set to be 50. The dimensions of the image vectors may be 4 x 64, which is not limiting to the present application.
Finally, the image vector is input to the image decoding layer 403 to obtain a picture output, and the picture output is amplified to a complete image to obtain a final poster picture.
In some embodiments, the user may determine whether the generated poster image meets the business requirement, and if not, may re-adjust the input text to re-generate the poster image; if yes, the poster picture can be saved in a server of a local or poster generation system, and basic materials are provided for poster generation.
In other embodiments, in the car rental field, after obtaining the poster image, the user of the poster generation system may save the poster image locally, and the intermediary agent of the car rental company may perform secondary development on the poster image through the interface for requesting the WeChat, and replace the corresponding content in the poster with the head portrait, the job number, the nickname, etc. of the intermediary agent.
Fig. 5 shows a flowchart of a method for generating a poster according to another embodiment provided in the present application, and as can be seen from fig. 5, the present embodiment includes the following steps S501 to S513:
step S501: and obtaining a graphic design file, and taking the graphic design file as a poster template. In the step, the poster template can also be generated by obtaining a plurality of configuration elements of the poster template and typesetting the plurality of configuration elements.
Step S502: judging whether the graphic design file is a PSD format file, if so, turning to step S504; otherwise, step S503 is shifted.
Step S503: and extracting target prompt information such as head description, indirect objects, cross index tables, file tail and the like in the graphic design file according to the data storage format of the graphic design file.
Step S504: and extracting target prompt information such as file header, detailed color mode information, image source data, image layer and mask data, image information and the like of the graphic design file according to the data storage format of the graphic design file.
Step S505: and executing word recognition, layer source feature matching and global picture feature matching on each item of target prompt information, and generating a template prompt text comprising a global prompt text, layer source prompt text and literal prompt text.
Step S506: and responding to the input operation of the user on the front-end interface, and acquiring the requirement information.
Step S507: and extracting features of the demand information, and selecting a plurality of tags from the tags recommended by the pre-constructed recommended tag library to serve as demand prompt texts. The recommended tag library is constructed according to the following method: obtaining a plurality of text samples, and carrying out sentence segmentation on each text sample to obtain a plurality of sub-sentences; performing word segmentation and part-of-speech tagging on each sub-sentence, and reserving a plurality of candidate words with appointed parts-of-speech; constructing a candidate keyword graph according to the reserved multiple candidate words so as to determine the weight of each candidate word; and (3) arranging the candidate words in a descending order according to the weight, intercepting a plurality of candidate words which are ranked in front as labels, and forming a recommended label library.
Step S508: and performing de-duplication processing on the template prompt text and the demand prompt text to obtain a de-duplicated prompt text.
Step S509: and sorting the de-duplicated prompt texts in a descending order according to the degree of association with the demand information to obtain the sorted prompt texts.
Step S510: and according to the sorting order, assigning a weight to each text in the sorted prompt texts so as to preprocess the template prompt texts and the demand prompt texts and obtain preprocessed data.
Step S511: and (3) encoding the preprocessed data based on a text encoding layer of the intelligent poster generation model to obtain text vectors.
Step S512: and an image information creation layer based on the intelligent poster generation model, and denoising the random noise data according to the text vector to obtain an image vector.
Step S513: and decoding the image vector based on the image decoding layer of the intelligent poster generation model to obtain a poster picture.
Fig. 6 shows a schematic structural diagram of a poster generating apparatus according to an embodiment provided herein, where the apparatus 600 is used for a server side 102 of a poster generating system; the apparatus 600 includes a first text generation module 601, a second text generation module 602, and a generation module 603, wherein:
The first text generation module 601 is configured to obtain a poster template, extract image information of the poster template, and generate a template prompt text according to the image information.
The second text generation module 602 is configured to obtain at least one item of requirement information according to a preset rule, and generate a requirement prompt text according to the requirement information, where the preset rule is used to describe a plurality of personalized parameters required for generating a poster.
And the generating module 603 is configured to input the template prompt text and the demand prompt text into a smart poster generating model for processing, so as to obtain a poster picture.
In some embodiments of the present application, in the foregoing apparatus, the first text generation module 601 is configured to obtain a graphic design file, and use the graphic design file as the poster template; or, obtaining a plurality of configuration elements of the poster template, typesetting the plurality of configuration elements to generate the poster template, wherein the configuration elements comprise basic elements and business elements, the basic elements comprise at least one of lines, geometric figures, pictures and texts, and the business elements comprise at least one of two-dimensional codes, applet codes and head portraits.
In some embodiments of the present application, in the above device, the poster template is a graphic design file; the first text generation module 601 is configured to extract target prompt information from the poster template according to a file type of the graphic design file; if the graphic design file is a first format file, extracting at least one of a file header, detailed color mode information, image source data, image layer and mask data and image information of the graphic design file according to a data storage format of the graphic design file; if the graphic design file is a file in a second format, extracting multi-level information from the graphic design file according to a data storage format of the graphic design file, wherein coarse-grained information in the multi-level information comprises at least one of beginning description, indirect objects, a cross index table and file tail; and executing at least one of character recognition, layer picture source feature matching and global picture feature matching on the target prompt information of each item to generate the template prompt text, wherein the template prompt text is one or more of global prompt text, layer picture source prompt text and literal prompt text.
In some embodiments of the present application, in the foregoing apparatus, in the preset rule, input forms of the plurality of personalization parameters are: the plurality of personalization parameters = [ picture subject ] (or [ adjective ] [ subject ]), [ detail setting ], [ modifier or artist ]; the second text generation module 602 is configured to display a front-end interface on the client, where the front-end interface is generated according to the preset rule, and the front-end interface includes a plurality of personalized parameter controls, and the plurality of personalized parameter controls are associated with a pre-built recommendation tag library; and responding to the input operation of the user on the front-end interface, and acquiring the requirement information.
In some embodiments of the present application, the apparatus further includes a tag library construction module, where the tag library construction module is configured to obtain a plurality of text samples, and segment each text sample into sentences to obtain a plurality of sub-sentences; performing word segmentation and part-of-speech tagging on each sub-sentence, and reserving a plurality of candidate words with appointed parts-of-speech; constructing a candidate keyword graph according to the reserved multiple candidate words so as to determine the weight of each candidate word; and according to the weight, descending order is carried out on the candidate words, and a plurality of candidate words which are ranked in front are intercepted to be used as labels, so that the recommended label library is formed.
In some embodiments of the present application, in the foregoing apparatus, the apparatus further includes a preprocessing module, where the preprocessing module is configured to perform a deduplication process on the template prompt text and the demand prompt text to obtain a deduplicated prompt text; according to the degree of association with the demand information, the de-duplicated prompt texts are ordered in a descending order to obtain ordered prompt texts; and according to the sorting order, assigning a weight to each text in the sorted prompt texts so as to preprocess the template prompt texts and the demand prompt texts.
In some embodiments of the present application, in the above apparatus, the poster intelligent generation model is constructed based on a Stable diffration model, and the poster intelligent generation model includes a text encoding layer, an image information creating layer, and an image decoding layer connected in sequence; the generating module 603 is configured to encode the template prompt text and the demand prompt text based on the text encoding layer to obtain a text vector; denoising random noise data according to the text vector based on the image information creation layer to obtain an image vector; and decoding the image vector based on the image decoding layer to obtain the poster picture.
It should be noted that, the generating device of any of the above-mentioned poster may be in one-to-one correspondence with the generating method of the poster, which is not described herein.
Fig. 7 shows a schematic structural diagram of an electronic device according to an embodiment of the present application. As shown in fig. 7, at the hardware level, the electronic device comprises a processor, optionally together with an internal bus, a network interface, a memory. The Memory may include a Memory, such as a Random-Access Memory (RAM), and may further include a non-volatile Memory (non-volatile Memory), such as at least 1 disk Memory. Of course, the electronic device may also include hardware required for other services.
The processor, network interface, and memory may be interconnected by an internal bus, which may be an ISA (Industry Standard Architecture ) bus, a PCI (Peripheral Component Interconnect, peripheral component interconnect standard) bus, or EISA (Extended Industry Standard Architecture ) bus, among others. The buses may be divided into address buses, data buses, control buses, etc. For ease of illustration, only one bi-directional arrow is shown in FIG. 7, but not only one bus or type of bus.
And the memory is used for storing programs. In particular, the program may include program code including computer-operating instructions. The memory may include memory and non-volatile storage and provide instructions and data to the processor.
The processor reads the corresponding computer program from the nonvolatile memory to the memory and then runs the computer program to form the generating device of the poster on the logic level. And the processor is used for executing the program stored in the memory and particularly used for executing the method.
The processor may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware in a processor or by instructions in the form of software. The processor may be a general-purpose processor, including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), etc.; but also digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), field programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The disclosed methods, steps, and logic blocks in the embodiments of the present application may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of a method disclosed in connection with the embodiments of the present application may be embodied directly in hardware, in a decoded processor, or in a combination of hardware and software modules in a decoded processor. The software modules may be located in a random access memory, flash memory, read only memory, programmable read only memory, or electrically erasable programmable memory, registers, etc. as well known in the art. The storage medium is located in a memory, and the processor reads the information in the memory and, in combination with its hardware, performs the steps of the above method.
The electronic device may execute the method for generating the poster provided in the embodiments of the present application and implement the function of the generating device for the poster in the embodiment shown in fig. 6, which is not described herein again.
The embodiments of the present application also provide a computer-readable storage medium storing one or more programs, the one or more programs including instructions, which when executed by an electronic device including a plurality of application programs, enable the electronic device to perform the method for generating a poster provided by the embodiments of the present application.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of computer-readable media.
Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other identical elements in a process, method, article or apparatus that comprises the element.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and changes may be made to the present application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. which are within the spirit and principles of the present application are intended to be included within the scope of the claims of the present application.
Claims (10)
1. The method is characterized by being applied to a server side of a poster generation system; the method comprises the following steps:
acquiring a poster template, extracting image information of the poster template, and generating a template prompt text according to the image information;
acquiring at least one item of demand information according to a preset rule, and generating a demand prompt text according to the demand information, wherein the preset rule is used for describing a plurality of personalized parameters required for generating a poster;
And inputting the template prompt text and the demand prompt text into a poster intelligent generation model for processing to obtain a poster picture.
2. The method of claim 1, wherein the obtaining a poster template comprises:
obtaining a graphic design file, and taking the graphic design file as the poster template;
or alternatively, the process may be performed,
obtaining a plurality of configuration elements of the poster template, typesetting the plurality of configuration elements to generate the poster template, wherein the configuration elements comprise basic elements and business elements, the basic elements comprise at least one of lines, geometric figures, pictures and texts, and the business elements comprise at least one of two-dimensional codes, applet codes and head portraits.
3. The method of claim 1, wherein the poster template is a graphic design file;
the extracting the image information of the poster template, generating a template prompt text according to the image information, comprises the following steps:
extracting target prompt information from the poster template according to the file type of the graphic design file; if the graphic design file is a first format file, extracting at least one of a file header, detailed color mode information, image source data, image layer and mask data and image information of the graphic design file according to a data storage format of the graphic design file; if the graphic design file is a file in a second format, extracting multi-level information from the graphic design file according to a data storage format of the graphic design file, wherein coarse-grained information in the multi-level information comprises at least one of beginning description, indirect objects, a cross index table and file tail;
And executing at least one of character recognition, layer picture source feature matching and global picture feature matching on the target prompt information of each item to generate the template prompt text, wherein the template prompt text is one or more of global prompt text, layer picture source prompt text and literal prompt text.
4. The method according to claim 1, wherein in the preset rule, the input form of the plurality of personalization parameters is:
the plurality of personalization parameters = [ picture subject ] (or [ adjective ] [ subject ]), [ detail setting ], [ modifier or artist ];
the obtaining at least one piece of demand information according to a preset rule includes:
displaying a front-end interface on a client, wherein the front-end interface is generated according to the preset rule, the front-end interface comprises a plurality of personalized parameter controls, and the personalized parameter controls are associated with a pre-built recommended label library;
and responding to the input operation of the user on the front-end interface, and acquiring the requirement information.
5. The method of claim 4, wherein the library of recommended tags is constructed according to the method of:
obtaining a plurality of text samples, and carrying out sentence segmentation on each text sample to obtain a plurality of sub-sentences;
Performing word segmentation and part-of-speech tagging on each sub-sentence, and reserving a plurality of candidate words with appointed parts-of-speech;
constructing a candidate keyword graph according to the reserved multiple candidate words so as to determine the weight of each candidate word;
and according to the weight, descending order is carried out on the candidate words, and a plurality of candidate words which are ranked in front are intercepted to be used as labels, so that the recommended label library is formed.
6. The method of claim 1, further comprising, prior to the step of inputting the template prompt text and the demand prompt text into a poster intelligence generation model for processing:
performing de-duplication processing on the template prompt text and the required prompt text to obtain a de-duplicated prompt text;
according to the degree of association with the demand information, the de-duplicated prompt texts are ordered in a descending order to obtain ordered prompt texts;
and according to the sorting order, assigning a weight to each text in the sorted prompt texts so as to preprocess the template prompt texts and the demand prompt texts.
7. The method of claim 1, wherein the intelligent poster generation model is constructed based on a Stable diffration model, and comprises a text encoding layer, an image information creation layer and an image decoding layer which are connected in sequence;
Inputting the template prompt text and the demand prompt text into a poster intelligent generation model for processing to obtain a poster picture, wherein the method comprises the following steps of:
coding the template prompt text and the demand prompt text based on the text coding layer to obtain text vectors;
denoising random noise data according to the text vector based on the image information creation layer to obtain an image vector;
and decoding the image vector based on the image decoding layer to obtain the poster picture.
8. The device is used for a server side of a poster generation system; the device comprises:
the first text generation module is used for acquiring a poster template, extracting image information of the poster template and generating a template prompt text according to the image information;
the second text generation module is used for acquiring at least one item of demand information according to a preset rule and generating a demand prompt text according to the demand information, wherein the preset rule is used for describing a plurality of personalized parameters required for generating the poster;
and the generating module is used for inputting the template prompt text and the demand prompt text into a poster intelligent generating model for processing to obtain a poster picture.
9. A poster generation system comprising a client and a server in communication connection, wherein the server is deployed with the apparatus of claim 8;
the client is used for displaying a front-end interface, receiving the demand information through the front-end interface, sending the demand information to the server, receiving the poster image sent by the server and rendering.
10. A computer-readable storage medium storing one or more programs that, when executed by an electronic device comprising a plurality of application programs, cause the electronic device to perform the steps of the method of generating a poster as claimed in any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310349603.1A CN116385584A (en) | 2023-04-03 | 2023-04-03 | Poster generation method, device and system and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310349603.1A CN116385584A (en) | 2023-04-03 | 2023-04-03 | Poster generation method, device and system and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116385584A true CN116385584A (en) | 2023-07-04 |
Family
ID=86974533
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310349603.1A Pending CN116385584A (en) | 2023-04-03 | 2023-04-03 | Poster generation method, device and system and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116385584A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117171474A (en) * | 2023-09-18 | 2023-12-05 | 得效(上海)数字科技有限公司 | Multi-mode generation type AI content creation system, application and data flow direction method |
| CN117274429A (en) * | 2023-08-23 | 2023-12-22 | 北京安锐卓越信息技术股份有限公司 | Poster picture generation method and device, electronic equipment and storage medium |
-
2023
- 2023-04-03 CN CN202310349603.1A patent/CN116385584A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117274429A (en) * | 2023-08-23 | 2023-12-22 | 北京安锐卓越信息技术股份有限公司 | Poster picture generation method and device, electronic equipment and storage medium |
| CN117171474A (en) * | 2023-09-18 | 2023-12-05 | 得效(上海)数字科技有限公司 | Multi-mode generation type AI content creation system, application and data flow direction method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111488931B (en) | Article quality evaluation method, article recommendation method and corresponding devices | |
| CN116385584A (en) | Poster generation method, device and system and computer readable storage medium | |
| CN110705214B (en) | Automatic coding method and device | |
| CN113011186B (en) | Named entity recognition method, named entity recognition device, named entity recognition equipment and computer readable storage medium | |
| CN111914654B (en) | Text layout analysis method, device, equipment and medium | |
| CN113722438B (en) | Sentence vector generation method and device based on sentence vector model and computer equipment | |
| CN113762309B (en) | Object matching method, device and equipment | |
| CN112347787A (en) | Method, device and equipment for classifying aspect level emotion and readable storage medium | |
| CN104636428A (en) | Trademark recommendation method and device | |
| CN117058271A (en) | Method and computing device for generating commodity main graph background | |
| CN111985243A (en) | Emotion model training method, emotion analysis device and storage medium | |
| CN113222022A (en) | Webpage classification identification method and device | |
| CN114581710A (en) | Image recognition method, device, equipment, readable storage medium and program product | |
| CN113360654A (en) | Text classification method and device, electronic equipment and readable storage medium | |
| CN116469110A (en) | Image classification method, device, electronic equipment and computer readable storage medium | |
| CN117764669A (en) | Article recommendation method, device, equipment, medium and product | |
| CN113704508A (en) | Multimedia information identification method and device, electronic equipment and storage medium | |
| CN112395407A (en) | Method and device for extracting enterprise entity relationship and storage medium | |
| CN112312189A (en) | Video generation method and video generation system | |
| US20050050452A1 (en) | Systems and methods for generating an electronically publishable document | |
| CN114782720A (en) | Method, device, electronic device, medium, and program product for determining matching of document | |
| CN112785601B (en) | Image segmentation method, system, medium and electronic terminal | |
| CN116266259A (en) | Image and text structured output method and device, electronic equipment and storage medium | |
| CN113743118A (en) | Entity relation extraction method in legal document based on fusion relation information coding | |
| CN114398980A (en) | Cross-modal Hash model training method, encoding method, device and electronic equipment |
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 |