阅读说明：本技术 图片选择方法、系统、介质和电子设备 (Picture selection method, system, medium and electronic device ) 是由 刘春波 于 2018-06-01 设计创作，主要内容包括：本公开提供了一种图片选择方法,包括获取多个图片的图片数据,根据所述多个图片的图片数据确定每个图片的内容,其中,所述内容包括多个内容标签,根据每个图片的内容计算该图片的得分,以及根据多个图片中每个图片的得分,选择至少一个图片,其中,所述计算该图片的得分包括,获取所述图片的内容的至少一个内容标签以及与所述内容标签对应的置信值,以及根据各个所述内容标签的置信值,计算该图片的得分。本公开还提供了一种图片选择系统、一种电子设备以及一种计算机可读介质。(The disclosure provides a picture selection method, which includes acquiring picture data of a plurality of pictures, determining content of each picture according to the picture data of the plurality of pictures, wherein the content comprises a plurality of content tags, calculating a score of each picture according to the content of the picture, and selecting at least one picture according to the score of each picture in the plurality of pictures, wherein the calculating the score of the picture includes acquiring at least one content tag of the content of the picture and a confidence value corresponding to the content tag, and calculating the score of the picture according to the confidence value of each content tag. The present disclosure also provides a picture selection system, an electronic device, and a computer readable medium.)

1. A picture selection method, comprising:

acquiring picture data of a plurality of pictures;

determining the content of each picture according to the picture data of the pictures, wherein the content comprises a plurality of content tags;

calculating the score of each picture according to the content of the picture; and

Selecting at least one picture according to the score of each picture in the plurality of pictures,

Wherein the calculating the score of the picture comprises:

Acquiring at least one content tag of the content of the picture and a confidence value corresponding to the content tag; and

And calculating the score of the picture according to the confidence value of each content label.

2. The method of claim 1, wherein the obtaining at least one content tag of the content of the picture and a confidence value corresponding to the content tag comprises:

Identifying at least one content tag of the content in the picture through a convolutional neural network, and acquiring a confidence value corresponding to the content tag, wherein the identified content tag is ignored under the condition that the confidence value is lower than a preset threshold value.

3. The method of claim 1, wherein the calculating a score for the picture based on the confidence value of each of the content tags comprises:

Determining an interval to which the confidence value belongs;

Determining the grade of the content label according to the confidence value and the interval; and

based on the scores of the individual content tags, a score for the picture is calculated.

4. The method of claim 3, wherein the determining a score for the content tag based on the confidence value and the interval comprises:

judging whether the content tag is positive relevant content or negative relevant content to obtain a judgment result;

Determining the weight corresponding to the interval according to the judgment result; and

And determining the grade of the content label according to the confidence value and the weight.

5. The method of claim 4, further comprising:

determining a category to which a product corresponding to the picture belongs; and

And if the content label is positively correlated content, modifying the score of the content label based on the correlation between the content label and the category.

6. The method of claim 1, further comprising at least one of:

Determining a form of the picture, and prohibiting selection of the picture if the form of the picture does not comply with a predetermined rule, wherein the form of the picture comprises at least one of format, size or resolution; or

Identifying the content of the picture, and prohibiting the picture from being selected if the content of the picture does not comply with a predetermined rule.

7. A picture selection system, comprising:

The acquisition module is used for acquiring picture data of a plurality of pictures;

A first determining module, configured to determine content of each picture according to picture data of the multiple pictures, where the content includes multiple content tags;

The calculating module is used for calculating the score of each picture according to the content of the picture; and

A second determination module for selecting at least one picture according to the score of each of the plurality of pictures,

Wherein the calculation module comprises:

the obtaining submodule is used for obtaining at least one content label of the content of the picture and a confidence value corresponding to the content label; and

And the determining submodule is used for calculating the score of the picture according to the confidence value of each content label.

8. The system of claim 7, wherein the acquisition submodule comprises:

The acquiring unit is used for identifying at least one content tag of the content in the picture through a convolutional neural network and acquiring a confidence value corresponding to the content tag, wherein the identified content tag is ignored under the condition that the confidence value is lower than a preset threshold value.

9. The system of claim 7, wherein the determination submodule comprises:

A first determining unit, configured to determine an interval to which the confidence value belongs;

A second determining unit, configured to determine a score of the content tag according to the confidence value and the interval; and

And the third determining unit is used for calculating the score of the picture based on the scores of the content tags.

10. The system of claim 9, wherein the second determination unit comprises:

the judging subunit is used for judging whether the content tag is positive relevant content or negative relevant content to obtain a judgment result;

The first determining subunit is configured to determine, according to the determination result, a weight corresponding to the interval; and

And the second determining subunit is used for determining the score of the content label according to the confidence value and the weight.

11. The system of claim 10, wherein the second determining unit further comprises:

The third determining subunit is used for determining the category to which the product corresponding to the picture belongs; and

And a correcting subunit, configured to correct the score of the content tag based on the correlation between the content tag and the category when the content tag is positively correlated content.

12. the system of claim 7, further comprising at least one of:

A first prohibiting module, configured to determine a form of the picture, and prohibit selection of the picture if the form of the picture does not comply with a predetermined rule, where the form of the picture includes at least one of a format, a size, or a resolution; or

And the second forbidding module is used for identifying the content of the picture and forbidding to select the picture under the condition that the content of the picture does not accord with a preset rule.

13. an electronic device, comprising:

One or more processors;

A storage device for storing one or more programs,

Wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to perform the method of any of claims 1-6.

14. a computer readable storage medium having stored thereon executable instructions which, when executed by a processor, cause the processor to perform the method of any one of claims 1 to 6.

Technical Field

the present disclosure relates to the field of internet technologies, and in particular, to a method, a system, a medium, and an electronic device for selecting a picture.

Background

The development of internet information technology greatly enriches the life of people. For example, a user may purchase a desired product on the web. Users often compare different products by browsing detailed pictures of the products presented by the clients. A better material picture can lead a user to feel the high quality of the product, and the click rate and the conversion rate are improved, so that the sales volume of the product is influenced. The richness of the picture content can be used for measuring the quality of the material picture.

In the prior art, the content of the material picture is audited mainly by manpower, which consumes a large amount of manpower resources, and the auditing process is influenced by personal factors.

Disclosure of Invention

In view of the above, the present disclosure provides a picture selection method, system, medium, and electronic device.

one aspect of the present disclosure provides a picture selection method, including obtaining picture data of a plurality of pictures, determining content of each picture according to the picture data of the plurality of pictures, wherein the content includes a plurality of content tags, calculating a score of each picture according to the content of the picture, and selecting at least one picture according to the score of each picture of the plurality of pictures, wherein the calculating the score of the picture includes obtaining at least one content tag of the content of the picture and a confidence value corresponding to the content tag, and calculating the score of the picture according to the confidence value of each content tag.

According to an embodiment of the disclosure, acquiring at least one content tag of the content of the picture and a confidence value corresponding to the content tag includes identifying at least one content tag of the content in the picture through a convolutional neural network and acquiring a confidence value corresponding to the content tag, wherein the identified content tag is ignored if the confidence value is lower than a preset threshold.

according to the embodiment of the disclosure, calculating the score of the picture according to the confidence value of each content tag comprises determining an interval to which the confidence value belongs, determining the score of the content according to the confidence value and the interval, and determining the score of the picture based on the score of each content.

According to an embodiment of the present disclosure, the determining the score of the content tag according to the confidence value and the interval includes determining that the content tag is positive related content or negative related content, obtaining a determination result, determining a weight corresponding to the interval according to the determination result, and determining the score of the content tag according to the confidence value and the weight.

according to the embodiment of the disclosure, the method further comprises determining a category to which a product corresponding to the picture belongs, and modifying the score of the content tag based on the correlation between the content tag and the category when the content tag is positively correlated content.

according to an embodiment of the present disclosure, the method further comprises at least one of determining a form of the picture, and in case the form of the picture does not comply with a predetermined rule, prohibiting selection of the picture, wherein the form of the picture comprises at least one of a format, a size or a resolution, or identifying a content of the picture, and in case the content of the picture does not comply with the predetermined rule, prohibiting selection of the picture.

Another aspect of the present disclosure provides a picture selection system, including an obtaining module configured to obtain picture data of a plurality of pictures, a first determining module configured to determine content of each picture according to the picture data of the plurality of pictures, where the content includes a plurality of content tags, a calculating module configured to calculate a score of each picture according to the content of the picture, and a second determining module configured to select at least one picture according to the score of each picture of the plurality of pictures, where the calculating module includes an obtaining sub-module configured to obtain at least one content tag of the content of the picture and a confidence value corresponding to the content tag, and a determining sub-module configured to calculate the score of the picture according to the confidence value of each content tag.

According to an embodiment of the disclosure, the obtaining sub-module includes an obtaining unit, configured to identify at least one content tag of content in a picture through a convolutional neural network, and obtain a confidence value corresponding to the content tag, where the identified content is ignored when the confidence value is lower than a preset threshold.

According to an embodiment of the present disclosure, the determining submodule includes a first determining unit configured to determine a section to which the confidence value belongs, a second determining unit configured to determine a score of the content tag according to the confidence value and the section, and a third determining unit configured to calculate a score of the picture based on the score of each content tag.

according to an embodiment of the present disclosure, the second determining unit includes a determining subunit configured to determine that the content tag is positive related content or negative related content, and obtain a determination result, a first determining subunit configured to determine a weight corresponding to the interval according to the determination result, and a second determining subunit configured to determine a score of the content tag according to the confidence value and the weight.

according to an embodiment of the present disclosure, the second determining unit further includes a third determining subunit configured to determine a category to which a product corresponding to the picture belongs, and a modifying subunit configured to modify, if the content tag is positive-correlation content, a score of the content tag based on a correlation between the content tag and the category.

According to an embodiment of the present disclosure, the system further includes at least one of a first prohibition module for determining a form of the picture, and prohibiting selection of the picture if the form of the picture does not comply with a predetermined rule, wherein the form of the picture includes at least one of a format, a size, or a resolution, or a second prohibition module for identifying content of the picture, and prohibiting selection of the picture if the content of the picture does not comply with the predetermined rule.

Another aspect of the disclosure provides an electronic device comprising one or more processors, a storage to store one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to perform the method of any of the above.

another aspect of the present disclosure provides a readable storage medium having stored thereon executable instructions that, when executed by a processor, cause the processor to perform the method as described above.

Another aspect of the disclosure provides a computer program comprising computer executable instructions for implementing the method as described above when executed.

According to the embodiment of the disclosure, the problem that the examination of the richness of the picture content mainly depends on manpower, a large amount of human resources are consumed and the examination process is greatly influenced by personal factors can be solved at least partially, so that the richness of the advertisement picture content can be automatically identified, further, in the advertisement putting process, the picture advertisement can be automatically and better screened, the effect after picture screening is improved, the influence of putting manpower and human factors is reduced, and the efficiency is improved.

Drawings

The above and other objects, features and advantages of the present disclosure will become more apparent from the following description of embodiments of the present disclosure with reference to the accompanying drawings, in which:

Fig. 1 schematically shows an application scenario of a picture selection method according to an embodiment of the present disclosure;

FIG. 2A schematically illustrates a flow chart of a picture selection method according to an embodiment of the present disclosure;

Fig. 2B schematically illustrates a flow chart of determining a score for each picture according to an embodiment of the present disclosure;

FIG. 3 schematically illustrates a flow chart for determining a score for the picture based on confidence values of respective ones of the content, in accordance with an embodiment of the disclosure;

FIG. 4 schematically illustrates a flow chart for determining a score for the content based on the confidence value and the interval, according to an embodiment of the disclosure;

FIG. 5 schematically illustrates a flow chart for determining a score for the content based on the confidence value and the interval, according to another embodiment of the disclosure;

FIG. 6 schematically shows a flow diagram for determining a score for each picture according to another embodiment of the present disclosure;

FIG. 7A schematically illustrates a block diagram of a picture selection system according to an embodiment of the present disclosure;

FIG. 7B schematically shows a block diagram of a computing module according to an embodiment of the disclosure;

FIG. 8 schematically illustrates a block diagram of a determining sub-module according to an embodiment of the disclosure;

Fig. 9 schematically shows a block diagram of a second determination unit according to an embodiment of the present disclosure;

Fig. 10 schematically shows a block diagram of a second determination unit according to another embodiment of the present disclosure; and

FIG. 11 schematically shows a block diagram of an electronic device according to an embodiment of the disclosure.

Detailed Description

Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. It should be understood that the description is illustrative only and is not intended to limit the scope of the present disclosure. In the following detailed description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the disclosure. It may be evident, however, that one or more embodiments may be practiced without these specific details. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present disclosure.

the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The terms "comprises," "comprising," and the like, as used herein, specify the presence of stated features, steps, operations, and/or components, but do not preclude the presence or addition of one or more other features, steps, operations, or components.

All terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art unless otherwise defined. It is noted that the terms used herein should be interpreted as having a meaning that is consistent with the context of this specification and should not be interpreted in an idealized or overly formal sense.

Where a convention analogous to "at least one of A, B and C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B and C" would include but not be limited to systems that have a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.). Where a convention analogous to "A, B or at least one of C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B or C" would include but not be limited to systems that have a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.). It will be further understood by those within the art that virtually any disjunctive word and/or phrase presenting two or more alternative terms, whether in the description, claims, or drawings, should be understood to contemplate the possibilities of including one of the terms, either of the terms, or both terms. For example, the phrase "a or B" should be understood to include the possibility of "a" or "B", or "a and B".

The embodiment of the disclosure provides a picture selection method, which includes acquiring picture data of a plurality of pictures, determining content of each picture according to the picture data of the plurality of pictures, wherein the content includes a plurality of content tags, calculating a score of each picture according to the content of the picture, and selecting at least one picture according to the score of each picture of the plurality of pictures, wherein the calculating the score of the picture includes acquiring at least one content tag of the content of the picture and a confidence value corresponding to the content tag, and calculating the score of the picture according to the confidence value of each content tag.

fig. 1 schematically shows an application scenario of a picture selection method according to an embodiment of the present disclosure. It should be noted that fig. 1 is only an example of an application scenario in which the embodiments of the present disclosure may be applied to help those skilled in the art understand the technical content of the present disclosure, but does not mean that the embodiments of the present disclosure may not be applied to other devices, systems, environments or scenarios.

As shown in FIG. 1, the application scenario includes a product 100 and advertisement material pictures 101-106 related to the product 100. When advertising is performed on the product 100, it is often necessary to select a better advertisement material picture from the advertisement material pictures 101 to 106. A good advertisement material picture can attract more users to click and display the effect. The richness of the picture content can be used for measuring the quality of the picture content, and in the advertisement industry, a better advertisement material picture can enable a user to feel the quality of a product, so that the click rate and the conversion rate are improved, and the advertisement putting effect is influenced.

However, it is very difficult to select one or more preferred advertisement material pictures from among a large number of advertisement materials. In the prior art, better advertisement material pictures are often manually screened from a plurality of advertisement material pictures, which needs to consume a large amount of human resources, and the auditing process is influenced by personal preference.

Therefore, the picture selection method is provided, so that a large amount of human resources are not needed for screening the advertisement material pictures, the screening effect is better, and the influence of human factors is avoided. Embodiments of the disclosed embodiments are described below with reference to fig. 2A, 2B, and 3 to 6.

fig. 2A schematically shows a flow chart of a picture selection method according to an embodiment of the present disclosure.

As shown in fig. 2A, the method includes operations S210 to S240.

In operation S210, picture data of a plurality of pictures is acquired. According to the embodiment of the present disclosure, the plurality of pictures may be, for example, pictures of appearances of the commodities or pictures of details of the commodities. For example, in the scenario shown in FIG. 1, the plurality of pictures acquired are advertising material pictures 101-106. According to an embodiment of the present disclosure, the picture data is, for example, pixel gray values of the picture, a size of the picture, and the like.

in operation S220, content of each picture is determined according to the picture data of the plurality of pictures, wherein the content includes a plurality of content tags. According to an embodiment of the present disclosure, for example, in the scenario shown in fig. 1, the content of the determined picture 103 may include a content tag 113 and a content tag 123, where the content tag 113 is milk and the content tag 123 is text.

In operation S230, a score of each picture is calculated according to the content of the picture. The calculating the score of the picture comprises the steps of obtaining at least one content label of the content of the picture and a confidence value corresponding to the content label, and calculating the score of the picture according to the confidence value of each content label.

In operation S240, at least one picture is selected according to the score of each of the plurality of pictures. According to the embodiment of the disclosure, for example, in the scenario shown in fig. 1, according to the scores of the pictures 101 to 106, for example, the first 1 to 3 pictures with higher scores can be selected as the pictures for advertisement delivery.

Fig. 2B schematically shows a flowchart of calculating a score of the picture in operation S230 according to an embodiment of the present disclosure.

As shown in fig. 2B, the method includes operations S231 and S232.

In operation S231, at least one content tag of the content of the picture and a confidence value corresponding to the content tag are acquired.

According to an embodiment of the present disclosure, for example, in the scenario shown in fig. 1, it is recognized that the content tag 113 in the picture 103 is milk with a confidence value of 0.9. Similarly, other content tags in picture 103 and content tags in other pictures in the scene shown in fig. 1 may be identified, as well as confidence values corresponding to the content tags.

According to the embodiment of the disclosure, acquiring at least one content tag of the content of the picture and a confidence value corresponding to the content tag comprises identifying at least one content tag in the content of the picture through a convolutional neural network and acquiring a confidence value corresponding to the content tag, wherein the identified content is ignored when the confidence value is lower than a preset threshold value. For example, in the case that the preset threshold is 0.65, the content tag 113 in the picture 103 is identified as milk through the convolutional neural network, and if the confidence value is less than 0.65, for example, 0.6, the identified content tag 113 is ignored.

In operation S232, a score of the picture is calculated according to the confidence value of each of the content tags. According to an embodiment of the present disclosure, the score of the picture may be a result of an accumulation of scores of the respective content tags in the picture.

Fig. 3 schematically shows a flowchart of calculating a score of the picture according to the confidence value of each of the content tags in operation S232 according to an embodiment of the present disclosure.

As shown in fig. 3, the method includes operations S310 to S330.

In operation S310, an interval to which the confidence value belongs is determined. According to an embodiment of the present disclosure, for example, the interval of the confidence value is divided into [1.0, 0.9 ], [0.9, 0.8 ], [0.8, 0.7), [0.7, 0.65), and if the confidence value is, for example, 0.9, the interval to which the confidence value 0.9 belongs is [0.9, 0.8).

In operation S320, a score of the content is determined according to the confidence value and the interval.

next, operation S320 of the embodiment of the present disclosure is described with reference to fig. 4 and 5.

Fig. 4 schematically shows a flowchart for determining a score of the content tag according to the confidence value and the interval at operation S320 according to an embodiment of the present disclosure.

as shown in fig. 4, the method includes operations S410 to S430.

In operation S410, it is determined whether the content tag is positive related content or negative related content, and a determination result is obtained. According to the embodiment of the disclosure, the content tags in the picture may be positive related content or negative related content, the positive related content may be content which makes the advertising effect of the picture better, and the negative related content may be content which makes the advertising effect of the picture worse or does not bring the advertising effect. For example, the words, lines, tables, etc. in the picture may have a poor or no advertising effect, while the picture in the picture makes the picture have a good advertising effect. In the scenario shown in fig. 1, the content tag 113 in the picture 103 may be positive related content and the content tag 123 may be negative related content.

In operation S420, a weight corresponding to the interval is determined according to the determination result. According to the embodiment of the present disclosure, different intervals correspond to different weights, for example, the weighted values corresponding to the intervals in the scenario described in operation S310 are 1.5, 1.2, 0.8, and 0.6, respectively.

according to the embodiment of the present disclosure, the positive correlation content and the negative correlation content may be set with different weights, for example, the confidence value interval is [1.0, 0.9 ], [0.9, 0.8 ], [0.8, 0.7 ], [0.7, 0.65), the weight values corresponding to the positive correlation content are 1.5, 1.2, 0.8, and 0.6, respectively, and the weight values corresponding to the negative correlation content are 1.2, 1.0, 0.8, and 0.6, respectively.

In operation S430, a score of the content tag is determined according to the confidence value and the weight. According to an embodiment of the present disclosure, for example, in the scenario shown in fig. 1, the confidence value of the content tag 113 in the picture 103 is 0.9, and for positive correlation content, the weight of the interval [0.9, 0.8) corresponding to 0.9 is 1.2, and the score of the content tag 113 may be 1.2 × 0.9. The confidence value of the content tag 123 in the picture 110 is, for example, 0.8, the content tag 123 is, for example, negative-related content, and in the negative-related interval, the interval corresponding to the confidence value of 0.8 is, for example, [0.8, 0.7), the weight of the interval is, for example, 0.8, and the score of the content tag 123 may be- (0.8 × 0.8).

The method can obtain a fair and reasonable evaluation result by distinguishing the positive correlation content from the negative correlation content.

Fig. 5 schematically shows a flowchart for determining a score of the content tag according to the confidence value and the interval at operation S320 according to another embodiment of the present disclosure.

As shown in fig. 5, the method further includes operations S510 and S520 on the basis of the foregoing embodiment.

in operation S510, a category to which a product corresponding to the picture belongs is determined. According to the embodiment of the disclosure, for example, in the scenario shown in fig. 1, the products corresponding to the pictures 101 to 106 are the products 100, and the category to which the products 100 belong is, for example, beverages.

in operation S520, in the case that the content tag is positively correlated content, the score of the content tag is corrected based on the correlation between the content tag and the category. According to an embodiment of the present disclosure, for example, in the scenario shown in fig. 1, the content tag 113 is positively correlated content. The rating of the content is modified based on the relevance of the content tag 113 to the category, e.g., beverage. For example, if the content tag 113 is identified as milk, the content tag 113 has a strong correlation with the category beverage, and the score of the content tag 113 is modified, for example, the content tag 113 is weighted, or a modification factor is added to the score of the content tag 113.

according to the embodiment of the disclosure, the relevance of the content and the category can be determined by calculating the cosine similarity of the content and the category under the vector representation, so as to correct the score of the content.

referring back to fig. 3, in operation S330, a score of the picture is determined based on the scores of the respective content tags. According to an embodiment of the present disclosure, the score of the picture may be an accumulated result of scores of respective contents in the picture. For example, in the scenario shown in fig. 1, the picture 103 includes a content tag 113 and a content tag 123, and the score of the picture 103 may be an accumulated result of the scores of the content tag 113 and the content tag 123.

The method can automatically identify the abundance of the advertisement picture content, and further can automatically and better screen the picture advertisement in the advertisement putting process, improve the effect of the picture after screening, reduce the influence of putting manpower and human factors, and improve the efficiency.

according to an embodiment of the present disclosure, before operation S220, determining a form of the picture, wherein the form of the picture includes at least one of a format, a size, or a resolution, and prohibiting selection of the picture if the form of the picture does not comply with a predetermined rule. According to an embodiment of the present disclosure, for example, the predetermined rule may be that the resolution is greater than 1024 × 1024, and in the case where the resolution of a picture is less than 1024 × 1024, the picture is prohibited from being selected.

According to an embodiment of the present disclosure, identifying the content of the picture may be further included before operation S220, and in case the content of the picture does not comply with a predetermined rule, prohibiting the picture from being selected. According to an embodiment of the present disclosure, for example, the picture content contains sensitive words or the like, the selection of the picture is prohibited.

fig. 6 schematically shows a flow chart for determining a score for each picture according to another embodiment of the present disclosure.

as shown in fig. 6, the method includes operations S601 to S623.

In operation S601, at least one content tag in the content of the picture and a confidence value corresponding to the content tag are acquired, similar to operation S231 described above. For example, in the scenario shown in fig. 1, at least one content in picture 103 includes content tag 113 and content tag 123.

In operation S602, it is determined whether the confidence value of at least one content tag in the picture is greater than a threshold, for example, the threshold may be 0.65. When the confidence value of the at least one content tag is greater than 0.65, operation S603 is performed, and if not greater than 0.65, the content is discarded.

in operation S603, it is determined whether the content tag is negatively correlated, similar to operation S410 described above. If negative correlation is true, operation S614 is performed. Otherwise, operation S604 is performed.

In operation S614, a confidence value a of the negatively correlated content tag is determined.

For example, the confidence value intervals are [1.0, 0.9), [0.9, 0.8), [0.8, 0.7), [0.7, 0.65), respectively. Operations S615 to S617, similar to operation S310 described above, determine the interval to which the confidence value belongs. Operations S618 to S621, similar to operation S430 described above, determine a score of the content tag according to the confidence value and the weight. In operation S615, it is determined whether the confidence value a is greater than 0.9. If a > 0.9, operation S618 is performed to determine the score of the negatively correlated content tag, wherein the content tag score may be determined by the product n1 × a of the weight n1 corresponding to the interval [1.0, 0.9) to which the confidence value a belongs and the confidence value a. If a is less than 0.9, operation S616 is performed to determine whether the confidence value a is greater than 0.8. If a > 0.8, operation S619 is performed to determine the score n2 × a of the negatively correlated content tag, where n2 is the weight corresponding to the interval [0.9, 0.8). If a is less than 0.8, operation S617 is performed to determine whether the confidence value a is greater than 0.7. If a > 0.7, operation S620 is performed to determine the score n3 × a of the negatively correlated content tag, where n3 is the weight corresponding to the interval [0.8, 0.7). If a is less than 0.7, operation S621 is performed to determine the score n4 × a of the negative-relative content tag, where n4 is the weight corresponding to the interval [0.7, 0.65).

In operation S604, a confidence value b of the positively correlated content tag is determined.

Operations S606 to S608, similar to operation S310 described above, determine the interval to which the confidence value belongs. Operations S609 to S612, similar to operation S430 described above, determine a score of the content tag according to the confidence value and the weight. In operation S606, it is determined whether the confidence value b is greater than 0.9. If b > 0.9, operation S609 is performed to determine the score of the negatively correlated content tag, wherein the content tag score may be determined by the product p1 × b of the weight p1 corresponding to the interval [1.0, 0.9) to which the confidence value b belongs and the confidence value a. If b is less than 0.9, operation S607 is executed to determine whether the confidence value b is greater than 0.8. If b > 0.8, operation S610 is performed to determine the score p2 × b of the negatively correlated content tag, where p2 is the weight corresponding to the interval [0.9, 0.8). If b is less than 0.8, operation S608 is performed to determine whether the confidence value b is greater than 0.7. If b > 0.7, operation S611 is performed to determine the score p3 × b of the negatively correlated content tag, where p3 is the weight corresponding to the interval [0.8, 0.7). If b < 0.7, operation S612 is performed to determine a score p4 × b of the negative relevant content tag, where p4 is a weight corresponding to the interval [0.7, 0.65).

If the content of the picture is related to the category of the product, in operation S605, the score of the content tag is corrected based on the correlation between the content tag and the category, similar to operation S520 described above, in the case that the content tag is positively related content. For example, the weight of the content tag is increased, or a correction factor is added to the score of the content tag, etc.

Operations S622, S613, and S623 determine scores of the pictures based on the scores of the respective content tags, similar to operation S330 described above. In operation S622, the negative correlation value is counted. For example, scores of negatively correlated content tags in the content of the picture are accumulated to determine a negative correlation value. In operation S613, positive correlation values are counted. For example, scores of positively correlated content tags in the content of the picture are accumulated to determine a positive correlation value. In operation S623, a score is output according to the positive correlation value and the negative correlation value. For example, the positive correlation value and the negative correlation value are added to obtain a final score.

fig. 7A schematically illustrates a block diagram of a picture selection system 700 according to an embodiment of the present disclosure.

as shown in fig. 7A, the picture selection system 700 includes an acquisition module 710, a first determination module 720, a calculation module 730, and a second determination module 740.

The obtaining module 710, for example, performs operation S210 described above with reference to fig. 2A, for obtaining picture data of a plurality of pictures. According to an embodiment of the present disclosure, for example, in the scenario shown in FIG. 1, the plurality of pictures acquired are advertisement material pictures 101-106.

The first determining module 720, for example, performs operation S220 described above with reference to fig. 2A, for determining content of each picture according to the picture data of the plurality of pictures, wherein the content includes a plurality of content tags. According to an embodiment of the present disclosure, for example, in the scenario shown in fig. 1, the content of the determined picture 103 may include a content tag 113 and a content tag 123, where the content tag 113 is milk and the content tag 123 is text.

The calculating module 730, for example, performs the operation S230 described above with reference to fig. 2A, for calculating the score of each picture according to the content of the picture.

the second determining module 740, for example, performs operation S240 described above with reference to fig. 2A, for selecting at least one picture according to the score of each picture of the plurality of pictures. According to the embodiment of the disclosure, for example, in the scenario shown in fig. 1, according to the scores of the pictures 101 to 106, for example, the first 1 to 3 pictures with higher scores can be selected as the pictures for advertisement delivery.

fig. 7B schematically illustrates a block diagram of a computation module 730 according to an embodiment of the disclosure.

As shown in fig. 7B, the calculation module 730 includes an acquisition sub-module 731 and a determination sub-module 732.

The obtaining sub-module 731, for example, performs the operation S231 described above with reference to fig. 2B, for obtaining at least one content tag of the content of the picture and a confidence value corresponding to the content. According to an embodiment of the present disclosure, the obtaining sub-module 731 includes an obtaining unit, configured to identify at least one content tag of content in the picture through a convolutional neural network, and obtain a confidence value corresponding to the content tag, where in a case that the confidence value is lower than a preset threshold, the identified content is ignored.

the determining sub-module 732, for example, performs the operation S232 described above with reference to fig. 2B, for determining the score of the picture according to the confidence value of each of the content tags. According to an embodiment of the present disclosure, the score of the picture may be a result of an accumulation of scores of the respective content tags in the picture.

Fig. 8 schematically illustrates a block diagram of determining a sub-module 732 in accordance with an embodiment of the present disclosure.

As shown in fig. 8, the determination submodule 732 includes a first determination unit 810, a second determination unit 820, and a third determination unit 830.

the first determining unit 810, for example, performs the operation S310 described above with reference to fig. 3, for determining the interval to which the confidence value belongs. According to an embodiment of the present disclosure, for example, the interval of the confidence value is divided into [1.0, 0.9 ], [0.9, 0.8 ], [0.8, 0.7 ], [0.7, 0.65), the confidence value is for example 0.9, and the interval to which the confidence value 0.9 belongs is [0.9, 0.8).

the second determining unit 820, for example, performs the operation S320 described above with reference to fig. 3, and is configured to determine the score of the content tag according to the confidence value and the interval.

The third determining unit 830, for example, performs operation S330 described above with reference to fig. 3, for calculating a score of the picture based on the scores of the respective content tags. According to an embodiment of the present disclosure, the score of the picture may be an accumulated result of scores of respective content tags in the picture. For example, in the scenario shown in fig. 1, the picture 103 includes a content tag 113 and a content tag 123, and the score of the picture 103 may be an accumulated result of the scores of the content tag 113 and the content tag 123.

Fig. 9 schematically shows a block diagram of the second determination unit 820 according to an embodiment of the present disclosure.

As shown in fig. 9, the second determining unit 820 includes a judging sub-unit 910, a first determining sub-unit 920, and a second determining sub-unit 930.

The determining subunit 910, for example, executes the operation S410 described above with reference to fig. 4, and is configured to determine that the content tag is positive related content or negative related content, and obtain a determination result. According to the embodiment of the disclosure, the content tags in the picture may be positive related content or negative related content, the positive related content may be content which makes the advertising effect of the picture better, and the negative related content may be content which makes the advertising effect of the picture worse or does not bring the advertising effect. For example, the words, lines, tables, etc. in the picture may have a poor or no advertising effect, while the picture in the picture makes the picture have a good advertising effect. In the scenario shown in fig. 1, the content tag 113 in the picture 103 may be positive related content and the content tag 123 may be negative related content.

the first determining subunit 920, for example, performs the operation S420 described above with reference to fig. 4, and is configured to determine the weight corresponding to the interval according to the determination result. According to the embodiment of the present disclosure, different intervals correspond to different weights, for example, the weighted values corresponding to the intervals in the scenario described in operation S310 are 1.5, 1.2, 0.8, and 0.6, respectively.

A second determining subunit 930, for example performing operation S430 described above with reference to fig. 4, is configured to determine a score of the content tag according to the confidence value and the weight. According to an embodiment of the present disclosure, for example, in the scenario shown in fig. 1, the confidence value of the content tag 113 in the picture 103 is 0.9, and for positive correlation content, the weight of the interval [0.9, 0.8) corresponding to 0.9 is 1.2, and the score of the content tag 113 may be 1.2 × 0.9. The confidence value of the content tag 123 in the picture 103 is, for example, 0.8, the content tag 113 is, for example, negative-related content, and in the negative-related interval, the interval corresponding to the confidence value of 0.8 is, for example, [0.8, 0.7), the weight of the interval is, for example, 0.8, and the score of the content tag 123 may be- (0.8 × 0.8).

Fig. 10 schematically shows a block diagram of the second determination unit 820 according to another embodiment of the present disclosure.

As shown in fig. 10, the method further includes a third determining subunit 1010 and a modifying subunit 1020 on the basis of the foregoing embodiment.

The third determining subunit 1010, for example, performs operation S510 described above with reference to fig. 5, for determining a category to which a product corresponding to the picture belongs. According to the embodiment of the disclosure, for example, in the scenario shown in fig. 1, the products corresponding to the pictures 101 to 106 are the products 100, and the category to which the products 100 belong is, for example, beverages.

The modification subunit 1020, for example, executes the operation S520 described above with reference to fig. 5, for modifying the score of the content based on the relevance of the content tag to the category when the content tag is positively relevant content. According to an embodiment of the present disclosure, for example, in the scenario shown in fig. 1, the content tag 113 is positively correlated content. The score of the content tag is modified based on the relevance of the content tag 113 to the category, e.g., beverage. For example, if the content tag 113 is identified as milk, the content tag 113 has a strong correlation with the category beverage, and the score of the content tag 113 is modified, for example, the content tag 113 is weighted, or a modification factor is added to the score of the content tag 113.

According to an embodiment of the present disclosure, the picture selection system may further include a first prohibition module configured to determine a form of the picture, and prohibit the picture from being selected if the form of the picture does not comply with a predetermined rule, wherein the form of the picture includes at least one of a format, a size, or a resolution. According to an embodiment of the present disclosure, for example, the predetermined rule may be that the resolution is greater than 1024 × 1024, and in the case where the resolution of a picture is less than 1024 × 1024, the picture is prohibited from being selected.

According to an embodiment of the present disclosure, the picture selection system may further include a second prohibition module configured to identify content of the picture and prohibit selection of the picture if the content of the picture does not comply with a predetermined rule. According to an embodiment of the present disclosure, for example, the picture content contains sensitive words or the like, the selection of the picture is prohibited.

any number of modules, sub-modules, units, sub-units, or at least part of the functionality of any number thereof according to embodiments of the present disclosure may be implemented in one module. Any one or more of the modules, sub-modules, units, and sub-units according to the embodiments of the present disclosure may be implemented by being split into a plurality of modules. Any one or more of the modules, sub-modules, units, sub-units according to embodiments of the present disclosure may be implemented at least in part as a hardware circuit, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a system on a chip, a system on a substrate, a system on a package, an Application Specific Integrated Circuit (ASIC), or may be implemented in any other reasonable manner of hardware or firmware by integrating or packaging a circuit, or in any one of or a suitable combination of software, hardware, and firmware implementations. Alternatively, one or more of the modules, sub-modules, units, sub-units according to embodiments of the disclosure may be at least partially implemented as a computer program module, which when executed may perform the corresponding functions.

For example, any plurality of the acquisition module 710, the first determination module 720, the calculation module 730, the second determination module 740, the acquisition sub-module 731, the determination sub-module 732, the first determination unit 810, the second determination unit 820, the third determination unit 830, the judgment sub-unit 910, the first determination sub-unit 920, the second determination sub-unit 930, the third determination sub-unit 1010, the modification sub-unit 1020, the first prohibition module, and the second prohibition module may be combined and implemented in one module, or any one of them may be split into a plurality of modules. Alternatively, at least part of the functionality of one or more of these modules may be combined with at least part of the functionality of the other modules and implemented in one module. According to an embodiment of the present disclosure, at least one of the obtaining module 710, the first determining module 720, the calculating module 730, the second determining module 740, the obtaining sub-module 731, the determining sub-module 732, the first determining unit 810, the second determining unit 820, the third determining unit 830, the judging sub-unit 910, the first determining sub-unit 920, the second determining sub-unit 930, the third determining sub-unit 1010, the modifying sub-unit 1020, the first inhibiting module, and the second inhibiting module may be at least partially implemented as a hardware circuit, such as Field Programmable Gate Arrays (FPGAs), Programmable Logic Arrays (PLAs), systems on a chip, systems on a substrate, systems on a package, Application Specific Integrated Circuits (ASICs), or may be implemented in hardware or firmware in any other reasonable way of integrating or packaging circuits, or in any one of three implementations, software, hardware and firmware, or in any suitable combination of any of them. Alternatively, at least one of the obtaining module 710, the first determining module 720, the calculating module 730, the second determining module 740, the obtaining sub-module 731, the determining sub-module 732, the first determining unit 810, the second determining unit 820, the third determining unit 830, the judging sub-unit 910, the first determining sub-unit 920, the second determining sub-unit 930, the third determining sub-unit 1010, the correcting sub-unit 1020, the first inhibiting module and the second inhibiting module may be at least partially implemented as a computer program module, which may perform a corresponding function when executed.

FIG. 11 schematically shows a block diagram of an electronic device according to an embodiment of the disclosure. The electronic device shown in fig. 11 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present disclosure.

As shown in fig. 11, an electronic device 1100 according to an embodiment of the present disclosure includes a processor 1101, which can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM)1102 or a program loaded from a storage section 1108 into a Random Access Memory (RAM) 1103. The processor 1101 may comprise, for example, a general purpose microprocessor (e.g., a CPU), an instruction set processor and/or associated chipset, and/or a special purpose microprocessor (e.g., an Application Specific Integrated Circuit (ASIC)), among others. The processor 1101 may also include on-board memory for caching purposes. The processor 1101 may comprise a single processing unit or a plurality of processing units for performing the different actions of the method flows according to the embodiments of the present disclosure.

The RAM 1103 stores various programs and data required for the program. The processor 1101, the ROM1102, and the RAM 1103 are connected to each other by a bus 1104. The processor 1101 performs various operations of the method flow according to the embodiments of the present disclosure by executing programs in the ROM1102 and/or the RAM 1103. It is noted that the programs may also be stored in one or more memories other than the ROM1102 and RAM 1103. The processor 1101 may also perform various operations of the method flows according to the embodiments of the present disclosure by executing programs stored in the one or more memories.

Electronic device 1100 may also include input/output (I/O) interface 1105, input/output (I/O) interface 1105 also connected to bus 1104, according to an embodiment of the disclosure. Electronic device 1100 may also include one or more of the following components connected to I/O interface 1105: an input portion 1106 including a keyboard, mouse, and the like; an output portion 1107 including a signal output unit such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and a speaker; a storage section 1108 including a hard disk and the like; and a communication section 1109 including a network interface card such as a LAN card, a modem, or the like. The communication section 1109 performs communication processing via a network such as the internet. A driver 1110 is also connected to the I/O interface 1105 as necessary. A removable medium 1111 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 1110 as necessary, so that a computer program read out therefrom is mounted into the storage section 1108 as necessary.

According to embodiments of the present disclosure, method flows according to embodiments of the present disclosure may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network through the communication portion 1109 and/or installed from the removable medium 1111. The computer program, when executed by the processor 1101, performs the above-described functions defined in the system of the embodiment of the present disclosure. The systems, devices, apparatuses, modules, units, etc. described above may be implemented by computer program modules according to embodiments of the present disclosure.

The present disclosure also provides a computer-readable medium, which may be embodied in the apparatus/device/system described in the above embodiments; or may exist separately and not be assembled into the device/apparatus/system. The computer readable medium carries one or more programs which, when executed, implement the method described above.

according to embodiments of the present disclosure, a computer readable medium may be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present disclosure, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In contrast, in the present disclosure, a computer-readable signal medium may include a propagated data signal with computer-readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wired, optical fiber cable, radio frequency signals, etc., or any suitable combination of the foregoing.

For example, according to an embodiment of the present disclosure, a computer-readable medium may include the ROM1102 and/or the RAM 1103 and/or one or more memories other than the ROM1102 and the RAM 1103 described above.

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

those skilled in the art will appreciate that various combinations and/or combinations of features recited in the various embodiments and/or claims of the present disclosure can be made, even if such combinations or combinations are not expressly recited in the present disclosure. In particular, various combinations and/or combinations of the features recited in the various embodiments and/or claims of the present disclosure may be made without departing from the spirit or teaching of the present disclosure. All such combinations and/or associations are within the scope of the present disclosure.

the embodiments of the present disclosure have been described above. However, these examples are for illustrative purposes only and are not intended to limit the scope of the present disclosure. Although the embodiments are described separately above, this does not mean that the measures in the embodiments cannot be used in advantageous combination. The scope of the disclosure is defined by the appended claims and equivalents thereof. Various alternatives and modifications can be devised by those skilled in the art without departing from the scope of the present disclosure, and such alternatives and modifications are intended to be within the scope of the present disclosure.