CN111382295A - Image search result sorting method and device - Google Patents
Image search result sorting method and device Download PDFInfo
- Publication number
- CN111382295A CN111382295A CN201811612143.2A CN201811612143A CN111382295A CN 111382295 A CN111382295 A CN 111382295A CN 201811612143 A CN201811612143 A CN 201811612143A CN 111382295 A CN111382295 A CN 111382295A
- Authority
- CN
- China
- Prior art keywords
- image
- images
- aesthetic
- image search
- search result
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000013145 classification model Methods 0.000 claims abstract description 31
- 238000013210 evaluation model Methods 0.000 claims abstract description 31
- 238000004458 analytical method Methods 0.000 claims description 43
- 230000006399 behavior Effects 0.000 claims description 10
- 230000000875 corresponding effect Effects 0.000 description 97
- 238000004891 communication Methods 0.000 description 10
- 238000011156 evaluation Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 230000003287 optical effect Effects 0.000 description 5
- 238000012163 sequencing technique Methods 0.000 description 4
- 230000005236 sound signal Effects 0.000 description 4
- 238000001514 detection method Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 239000000463 material Substances 0.000 description 3
- 230000001133 acceleration Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000001815 facial effect Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the application discloses a method and a device for ordering image search results, wherein when the image search results of keywords are obtained, a first image set which belongs to a first image category is identified from the images of the image search results according to an image classification model, and aesthetic scores corresponding to the images in the first image set are determined through an aesthetic evaluation model corresponding to the first image category. When images in the image search result are sorted, the sorted positions of the images in the first image set in the image search result are determined by combining the aesthetic scores corresponding to the images in the first image set. Because the aesthetic score of one category can reflect the aesthetic feeling of the image to the user in the category, the possibility that the user selects the image with better aesthetic feeling is relatively higher, the image with higher aesthetic score can be preferentially displayed to the user, the possibility that the image with higher aesthetic score meets the image searching requirement is higher, and the image searching experience is improved.
Description
Technical Field
The present application relates to the field of image search, and in particular, to a method and an apparatus for ranking image search results.
Background
In the conventional image search, images in image search results are generally sorted by a relevance evaluation method for image search results corresponding to keywords.
However, image search is different from other types of search, and an image with high relevance to a keyword is not necessarily an image meeting the user requirements, so that according to the existing relevance evaluation mode, the obtained sequencing result is difficult to meet the image search requirements of the user, and the user search experience is not high.
Therefore, it is an urgent problem to improve the ranking effect of image search results.
Disclosure of Invention
In order to solve the technical problem, the application provides a method and a device for sorting image search results, so that images with high aesthetic scores can be preferentially displayed to a user, and the image search experience of the user is improved.
The embodiment of the application discloses the following technical scheme:
in a first aspect, an embodiment of the present application provides a method for ranking image search results, where the method includes:
acquiring an image search result corresponding to the keyword, wherein the image search result comprises a plurality of images;
identifying a first set of images from the image search results according to an image classification model; images in the first set of images belong to a first image category;
determining an aesthetic score corresponding to each image in the first image set through an aesthetic evaluation model corresponding to the first image category;
when images in the image search result are sorted, the sorted positions of the images in the first image set in the image search result are determined by combining the aesthetic scores corresponding to the images in the first image set.
Optionally, the identifying a first image set from the image search results according to an image classification model includes:
identifying the first image set and the second image set from the image search results according to the image classification model; images in the second set of images belong to a second image category different from the first image category;
the method further comprises the following steps:
determining an aesthetic rating score corresponding to each image in the second image set through an aesthetic evaluation model corresponding to the second image category;
and when the images in the image search result are sorted, determining the sorting positions of the images in the second image set in the image search result by combining the aesthetic scores corresponding to the images in the second image set respectively.
Optionally, the first image category has a correlation with a type to which the keyword belongs.
Optionally, if the first image category is a person category, the method further includes:
determining face analysis scores corresponding to the images in the first image set according to a face analysis model;
determining quality scores corresponding to the images in the first image set according to the face analysis scores corresponding to the images in the first image set and the aesthetic scores corresponding to the images in the first image set;
the determining the ranking position of the images in the first image set in the image search result by combining the aesthetic scores corresponding to the images in the first image set respectively comprises:
and determining the sorting position of the images in the first image set in the image search result by combining the quality scores corresponding to the images in the first image set respectively.
Optionally, the image search result is a plurality of images which are obtained by searching according to the keyword and the correlation between the images and the keyword meets a preset condition.
Optionally, if the keyword is input by the target user, the method further includes:
determining the image click characteristics of the target user according to the image search behavior of the target user;
after the sorting result of the images in the image searching result is determined, adjusting the sorting result according to the image clicking feature so as to advance the sorting position of the images which accord with the image clicking feature in the image searching result;
and displaying the image search result according to the adjusted sorting result.
In a second aspect, an embodiment of the present application provides an apparatus for ranking image search results, where the apparatus includes an obtaining unit, an identifying unit, a determining unit, and a ranking unit:
the acquisition unit is used for acquiring an image search result corresponding to the keyword, and the image search result comprises a plurality of images;
the identification unit is used for identifying a first image set from the image search results according to an image classification model; images in the first set of images belong to a first image category;
the determining unit is used for determining the aesthetic scores corresponding to the images in the first image set respectively through the aesthetic evaluation models corresponding to the first image categories;
and the sorting unit is used for determining the sorting positions of the images in the first image set in the image search results by combining the aesthetic scores corresponding to the images in the first image set when the images in the image search results are sorted.
Optionally, the identifying unit is further configured to identify the first image set and the second image set from the image search result according to the image classification model; images in the second set of images belong to a second image category different from the first image category;
the determining unit is further used for determining the aesthetic scores corresponding to the images in the second image set respectively through the aesthetic evaluation models corresponding to the second image categories;
the sorting unit is further used for determining sorting positions of the images in the second image set in the image search results by combining the aesthetic scores corresponding to the images in the second image set when the images in the image search results are sorted.
Optionally, the first image category has a correlation with a type to which the keyword belongs.
Optionally, if the first image type is a person type, the apparatus further includes a face analysis score determining unit and a quality score determining unit:
the face analysis score determining unit is used for determining face analysis scores corresponding to the images in the first image set according to a face analysis model;
the quality score determining unit is used for determining quality scores corresponding to the images in the first image set according to the face analysis scores corresponding to the images in the first image set and the aesthetic scores corresponding to the images in the first image set;
the sorting unit is further configured to determine, in combination with the quality scores corresponding to the images in the first image set, sorting positions of the images in the first image set in the image search result.
Optionally, the image search result is a plurality of images which are obtained by searching according to the keyword and the correlation between the images and the keyword meets a preset condition.
Optionally, if the keyword is input by the target user, the apparatus further includes an image click feature determining unit, an adjusting unit, and a displaying unit:
the image click feature determining unit is used for determining the image click feature of the target user according to the image search behavior of the target user;
the adjusting unit is used for adjusting the sorting result according to the image clicking feature after the sorting result of the images in the image searching result is determined so as to advance the sorting position of the images which accord with the image clicking feature in the image searching result;
and the display unit is used for displaying the image search result according to the adjusted sorting result.
In a third aspect, an apparatus for ranking image search results is provided, including a memory, and one or more programs, where the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for:
acquiring an image search result corresponding to the keyword, wherein the image search result comprises a plurality of images;
identifying a first set of images from the image search results according to an image classification model; images in the first set of images belong to a first image category;
determining an aesthetic score corresponding to each image in the first image set through an aesthetic evaluation model corresponding to the first image category;
when images in the image search result are sorted, the sorted positions of the images in the first image set in the image search result are determined by combining the aesthetic scores corresponding to the images in the first image set.
In a fourth aspect, embodiments of the present application provide a machine-readable medium having stored thereon instructions, which when executed by one or more processors, cause an apparatus to perform a method of ranking image search results as described in one or more of the first aspects.
According to the technical scheme, when the image search result which corresponds to the keyword and comprises a plurality of images is obtained, the first image set which belongs to the first image category is identified from the images of the image search result according to the image classification model, and the aesthetic scores which correspond to the images in the first image set are determined through the aesthetic evaluation model which corresponds to the first image category. When images in the image search result are sorted, the sorted positions of the images in the first image set in the image search result are determined by combining the aesthetic scores corresponding to the images in the first image set. For the images belonging to an image category, the aesthetic scores of the category can reflect the aesthetic feelings of the images to the user under the category, the possibility that the user selects the images with better aesthetic feelings is higher, therefore, the ranking positions of the images with higher aesthetic scores are advanced, the ranking positions of the images with lower aesthetic scores are advanced, the images with higher aesthetic scores can be preferentially displayed to the user, the probability that the images with the higher aesthetic scores meet the image searching requirements of the user is higher, and the image searching experience of the user is improved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without inventive exercise.
Fig. 1 is a flowchart of a method for ranking image search results according to an embodiment of the present disclosure;
fig. 2 is a device structure diagram of an apparatus for sorting image search results according to an embodiment of the present application;
fig. 3 is a structural diagram of an apparatus for sorting image search results according to an embodiment of the present application;
fig. 4 is a block diagram of a server according to an embodiment of the present application.
Detailed Description
Embodiments of the present application are described below with reference to the accompanying drawings.
In the conventional image search, images in image search results are generally sorted by a relevance evaluation method for image search results corresponding to keywords. And the images with high relevance to the keywords are displayed to the user preferentially because the sequencing positions of the images are closer to the front.
However, images are different from search items such as files and pages, aesthetic feelings of the images directly influence whether a user wants to click to view the images, for example, the correlation between one image in an image search result and a keyword is very high, but the aesthetic feelings are not satisfactory, the possibility that the user clicks to view the image is not high, and therefore, the image with the high correlation with the keyword is not always an image meeting the requirements of the user, and the problem of improving the ranking effect of the image search result is a problem which needs to be solved urgently at present.
To this end, the embodiment of the present application provides a method for ranking image search results, which determines an aesthetic score of an image category in the image search results through an image classification model and an aesthetic evaluation model, and adjusts a ranking position of such an image in the image search results in combination with the aesthetic score. The sorting method may be applied to a processing device, which may be a terminal, a computer, a server, etc. The image classification model and the aesthetic evaluation model may be configured in one processing device, or may be configured in different processing devices.
For an image belonging to a certain image category, the aesthetic score of the category can reflect the aesthetic feeling of the image to a user under the category, the possibility that the user selects the image with better aesthetic feeling is relatively higher, so that the ranking position of the image with higher aesthetic score is advanced, the ranking position of the image with lower aesthetic score is advanced, the image with higher aesthetic score can be preferentially shown to the user, and the probability that the image with higher aesthetic score meets the image searching requirement of the user is higher, so that the image searching experience of the user is improved.
The method for sorting image search results provided by the embodiment of the application is described next with reference to fig. 1, and the method includes:
s101: and acquiring an image search result corresponding to the keyword.
The image search result is an image which is obtained by the search engine according to the keyword and is related to the keyword, and the image search result comprises a plurality of images.
The image classification model and the aesthetic evaluation model can determine the aesthetic scores of the images of a certain image category in the image search results, and adjust the ranking positions of the images in the image search results by combining the aesthetic scores. In order to improve the efficiency of adjusting the ranking positions in combination with the aesthetic scores, images for which the aesthetic scores need to be determined can be selected in a targeted manner.
In a possible implementation manner, the image search result is a plurality of images which are obtained according to the keyword search and the correlation between the images and the keyword meets a preset condition.
That is, in this possible implementation, the image search result may be a part of all images obtained by the keyword search, for example, the top N images with the highest relevance.
The relevance of the partial image and the keywords is relatively high, so that the ranking position obtained according to the relevance is relatively forward, the probability that the ranking position is forward after the aesthetic score is adjusted by determining the aesthetic score of the partial image and combining the aesthetic score is relatively high, the partial image is easier to be displayed to a user in a priority mode, and the probability that the user clicks and views the partial image is further improved.
S102: a first set of images is identified from the image search results according to an image classification model.
The image classification model may be pre-trained and at least performs the function of identifying an image of a certain image class from the images.
In the embodiment of the application, a first image set belonging to a first image category is identified from image search results according to an image classification model, and the first image set comprises at least one image in the image search results.
The embodiment of the application does not limit the dividing manner or the dividing granularity of the image categories, for example, the image categories may be divided according to different display objects in the image, and may include people, objects, materials, and scenery. The first image category may be any of the categories described above. That is, the embodiment of the present application may identify any image category in the image search result, and adjust the ranking position by the determined aesthetic score for that type of image.
In one possible implementation, the first image category has a correlation with a type to which the keyword belongs. That is, when image recognition is performed on the image search result corresponding to the keyword, an image with an image type correlated with the type of the keyword can be specifically recognized from the image search result through the image classification model.
The correlation described herein is understood to mean that the type to which the keyword belongs and the image type have an association relationship, such as similarity or identity. For example, the image types include four types, i.e., a human type, a real object type, a material type, and a landscape type, and if the keyword is "liu de hua" and the type to which the keyword belongs is determined to be a human, the human type in the image types has a correlation with the type "human" to which the keyword belongs, and the other three image types have no correlation with the type "human" to which the keyword belongs.
Therefore, in the implementation manner, the identified images in the first image set have correlation with the types of the keywords, so that the images in the first image set have better conformity with the image search requirements of the user for searching through the keywords, and therefore, the images are subjected to subsequent aesthetic scoring and the sequencing positions are adjusted, and the image search requirements of the user are more likely to be met.
It should be noted that, since the aesthetic evaluation model is determined according to the image category to perform the aesthetic scoring in the embodiment of the present application, different image categories should have different aesthetic evaluation manners, and the different aesthetic evaluation manners herein may refer to all or part of different aesthetic evaluation manners.
For example, in the images of people, the images with whiter, brighter skin color, more three-dimensional facial structure of people, better body proportion and the like can obtain higher aesthetic scores, and in the images of scenery, the images with higher color saturation and better composition proportion can obtain higher aesthetic scores. Therefore, when dividing the image categories, it is necessary to determine that the divided different image categories have different aesthetic evaluation manners in consideration of the above situation.
S103: and determining the aesthetic scores corresponding to the images in the first image set respectively through the aesthetic evaluation models corresponding to the first image categories.
The aesthetic evaluation model in this step scores the aesthetic orientation specifically for the images of the first image category. The aesthetic evaluation model may be pre-trained, for example, by a deep neural network.
The aesthetic evaluation model may score an aesthetic orientation based on the aesthetic feature corresponding to the first image category by analyzing features in the image that are associated with the aesthetic feature. The higher the aesthetic score of an image, the better the aesthetic feel of the image to the user.
A corresponding aesthetic score may be determined for the images in the first set of images via an aesthetic evaluation model corresponding to the first image category, wherein an image has a corresponding aesthetic score.
Alternatively, the aesthetic score may be a score of 0-1.
S104: when images in the image search result are sorted, the sorted positions of the images in the first image set in the image search result are determined by combining the aesthetic scores corresponding to the images in the first image set.
In this embodiment of the application, for the images in the first image set, the ranking positions of the images in the image search result may be determined only according to the aesthetic score, or the original ranking positions of the images in the first image set may be adjusted according to the aesthetic score on the basis of determining the ranking positions according to the high or low correlation between the images and the keywords in the image search result, for example, the ranking position of the image with the higher aesthetic score is advanced, and the ranking position of the image with the lower aesthetic score is advanced.
Regardless of the manner in which the ranking position of the images in the first image set in the image search results is determined in combination with the aesthetic score, the ranking position of at least some of the images with higher aesthetic scores can be advanced, and the top presentation position enables the user to view such images more quickly.
Therefore, when an image search result which corresponds to a keyword and comprises a plurality of images is obtained, a first image set which belongs to a first image category is identified from the images of the image search result according to an image classification model, and the aesthetic scores which correspond to the images in the first image set are determined through an aesthetic evaluation model which corresponds to the first image category. When images in the image search result are sorted, the sorted positions of the images in the first image set in the image search result are determined by combining the aesthetic scores corresponding to the images in the first image set. For the images belonging to an image category, the aesthetic scores of the category can reflect the aesthetic feelings of the images to the user under the category, the possibility that the user selects the images with better aesthetic feelings is higher, therefore, the ranking positions of the images with higher aesthetic scores are advanced, the ranking positions of the images with lower aesthetic scores are advanced, the images with higher aesthetic scores can be preferentially displayed to the user, the probability that the images with the higher aesthetic scores meet the image searching requirements of the user is higher, and the image searching experience of the user is improved.
In S102, an image belonging to a first image category may be identified from the image search results by the image classification model, and in addition to this, the image classification model may identify images of other image categories, for example, a second image category different from the first image category. That is, images of multiple image categories may be identified from the image search results by the image classification model, and corresponding image sets are formed.
Therefore, in one possible implementation manner, S102 may be:
identifying the first set of images and the second set of images from the image search results according to the image classification model.
Images in the second set of images belong to a second image category different from the first image category. For example, if the first image category is a person category and the second image category is a landscape category, then the images in the first image set are all images of the person category and the images in the second image category are all images of the landscape category.
It should be noted that some images may have multiple image categories, for example, an image a including a person in the field, and the image category may be both a person category and a landscape category, so when an image set of multiple different image categories, for example, a first image set and a second image set, is identified from an image search result by an image classification model, a certain image or certain images in the image search result may be in both the first image set and the second image set. For example, if the first image category is a person category and the second image category is a landscape category, the image a may be in the first image set or the second image set, in which case the image a may have an aesthetic score corresponding to the person category and an aesthetic score corresponding to the landscape category. When determining the ranking position of the image a in the image search result by combining the aesthetic scores, the ranking position of the image a can be determined by simultaneously considering the aesthetic scores of the image a corresponding to the people class and the aesthetic scores of the image a corresponding to the scenery class, for example, different weights are set to obtain a total aesthetic score by combining a plurality of aesthetic scores, and the ranking position of the image a is truncated by the total aesthetic score; or, the ranking position of the image a can be determined according to the higher score of the aesthetic score of the image a corresponding to the person class and the aesthetic score of the image a corresponding to the landscape class; alternatively, if a certain image type of the image a, for example, a person class, has a correlation with the type of the keyword, the aesthetic score of the person class may be used to determine the ranking position of the image a.
After the first image set and the second image set are determined, the images in the second image set need to be scored for aesthetic orientation. On the basis of the embodiment corresponding to fig. 1, the method further includes:
s201: and determining the aesthetic scores corresponding to the images in the second image set respectively through the aesthetic evaluation models corresponding to the second image categories.
The aesthetic evaluation model in this step scores the aesthetic orientation specifically for the images of the second image category. The aesthetic evaluation model may be pre-trained, for example, by a deep neural network.
The aesthetic evaluation model may score an aesthetic orientation based on the aesthetic feature corresponding to the second image category by analyzing features in the image that are associated with the aesthetic feature. The higher the aesthetic score of an image, the better the aesthetic feel of the image to the user.
A corresponding aesthetic score may be determined for the images in the second set of images via an aesthetic evaluation model corresponding to the second image category, wherein an image has a corresponding aesthetic score.
Alternatively, the aesthetic score may be a score of 0-1.
S202: and when the images in the image search result are sorted, determining the sorting positions of the images in the second image set in the image search result by combining the aesthetic scores corresponding to the images in the second image set respectively.
In this embodiment of the application, for the images in the second image set, the ranking positions of the images in the image search result may be determined only according to the aesthetic score, or the original ranking positions of the images in the second image set may be adjusted according to the aesthetic score on the basis of determining the ranking positions according to the high or low correlation between the images and the keywords in the image search result, for example, the ranking position of the image with the higher aesthetic score is advanced, and the ranking position of the image with the lower aesthetic score is advanced.
Regardless of the manner in which the ranking position of the images in the second image set in the image search results is determined in combination with the aesthetic score, the ranking position of at least some of the images with higher aesthetic scores can be advanced, and the top presentation position enables the user to view such images more quickly.
In summary, since the images of different image types are different greatly, the aesthetic feeling or appeal evaluation criteria of users for the images of different image types are also inconsistent, such as the aesthetic evaluation criteria from saturation or from image composition are greatly different for the material class and the character class. Therefore, in the embodiment of the application, when the images are scored in the aesthetic direction, different aesthetic scoring models can be adopted for scoring the images of different image types, different scoring standards and different scoring modes can be provided for different image types, the accuracy of aesthetic feeling which can be reflected by the aesthetic scoring is improved, and the image searching experience of a user can be improved by the determined image ranking position.
Aiming at the image with the image category of people, the embodiment of the application also provides a mode for further optimizing the sequencing position,
If the first image category is a person category, in a possible implementation manner, based on the embodiment corresponding to fig. 1, the method further includes:
s301: and determining the face analysis scores corresponding to the images in the first image set according to the face analysis model.
The face analysis model is used for analyzing the faces in the images, and can include face detection and recognition, face number analysis and the like. Through the above-mentioned face analysis process, face analysis scoring can be performed for the images in the first image set based on the analysis result. Wherein one image corresponds to one face analysis score.
It should be noted that the face analysis model may implement the face analysis function through one model, or may implement the face analysis function through multiple models, for example, the face analysis model may be divided into a face detection model and a face recognition model, where the face detection model may detect the position and number of faces in an image, and the face recognition model may recognize persons corresponding to the faces in the image.
For example, the keyword is 'song', the face analysis model can identify face information carried by each image in the image search result corresponding to the keyword, for example, whether a person in the image is a song, several persons and the like.
The face analysis score may reflect whether the face identified and detected in the image is associated with the character corresponding to the keyword, for example, in the image corresponding to the song in the above example, if one image only includes one face and is the song, the face analysis score is relatively high, and if one image includes a plurality of faces, only one face is the song, and the face analysis score is relatively low.
Through the description of the face analysis score in the previous section, the face analysis score can be made clear to reflect the degree of association between the face contained in the person image and the person corresponding to the keyword, or whether the image focuses on showing the person corresponding to the keyword.
The higher the degree of association between the face contained in one image and the task corresponding to the keyword, for example, if only the face of the person corresponding to the keyword is contained, and the face of another person is not contained, the higher the face analysis score of the image is.
Therefore, the possibility of whether the user selects the click selection can be influenced by the level of the face analysis score. The characters corresponding to the keywords can reflect the search requirements of the user, so that the comparability of the user to select the image with higher association degree with the characters corresponding to the keywords by clicking from the image search results is higher, that is, the higher the face analysis score of one image is, the higher the possibility that the user clicks to select the image is.
S302: and determining quality scores corresponding to the images in the first image set according to the face analysis scores corresponding to the images in the first image set and the aesthetic scores corresponding to the images in the first image set.
Because each image in the first image set has the corresponding aesthetic score and the face analysis score, the aesthetic score can reflect the aesthetic feeling of the image to the user, the face analysis score can reflect the degree of association between the face contained in the image and the person corresponding to the keyword, and both the two scores can be used for adjusting the ranking position of the image. For convenience of calculation, the two scores can be combined and synthesized to obtain the quality score corresponding to the image. One image in the first set of images has a corresponding quality score.
When the quality score is calculated by combining the aesthetic score and the face analysis score, the aesthetic score and the face analysis score may have different or the same weight coefficients, and how to set the weight coefficients is not limited in the embodiment of the present application, for example, the weight coefficients may be preset, or the correspondence may be adjusted according to different application scenarios.
After determining the quality score, in this embodiment, one possible implementation manner of S104 may be as shown in S303:
s303: and determining the sorting position of the images in the first image set in the image search result by combining the quality scores corresponding to the images in the first image set respectively.
In this embodiment of the application, for the images in the first image set, the ranking position of the images in the image search result may be determined only according to the quality score, or the original ranking position of the images in the first image set may be adjusted according to the quality score on the basis of determining the ranking position according to the relevance between the images and the keywords in the image search result, for example, the ranking position of the image with the higher quality score is advanced, and the ranking position of the image with the lower quality score is retarded.
No matter which way is combined with the quality scores to determine the ranking positions of the images in the first image set in the image search results, at least part of the images with higher quality scores can be advanced in ranking positions, and the front display position enables the user to view the images more quickly, so that the display effect of the image search results is improved.
Since the image searching behaviors of different users are different, the images selected by clicking of different users are different for the same keyword. Aiming at the phenomenon, the embodiment of the application provides a personalized sorting mode aiming at the image searching behavior of the user, and the image searching result corresponding to the user can be adjusted in a targeted manner by analyzing the image clicking characteristics of the user.
In a possible implementation manner, based on any one of the above embodiments, if the keyword is input by the target user, the method further includes:
s401: and determining the image click characteristics of the target user according to the image search behavior of the target user.
The image clicking characteristics of the target user can reflect the image clicking behavior characteristics of the target user during image searching, for example, the image clicking behavior characteristics show that the user is more interested in which type of image and is more prone to click selection.
Optionally, the image search behavior data of the target user may be modeled and analyzed by a deep learning method, and the tendency and the internal rule of image clicking of the target user are mainly learned. For example, through analysis, the image click features of the target user on the image are found to include relevance and freshness, i.e., the target user is more inclined to select the image with high relevance and freshness.
The types and the aforementioned image categories may be divided in the same manner or in different manners, and the present application is not limited thereto.
S402: after the ranking result of the images in the image search result is determined, the ranking result is adjusted according to the image clicking feature, so that the ranking position of the images which accord with the image clicking feature in the image search result is advanced.
S403: and displaying the image search result according to the adjusted sorting result.
It should be noted that the reference to "determining the ranking result of the images in the image search result" in S402 means that the ranking position of the images in the image set is determined in combination with the aesthetic score.
Determining an ordered position of the images in the first set of images, for example in combination with an aesthetic score; alternatively, the first and second electrodes may be,
for example, the ranking position of the images in the first set of images is determined in combination with the aesthetic score, and the ranking position of the images in the second set of images is determined in combination with the aesthetic score; alternatively, the first and second electrodes may be,
determining the ranking position of the images in the first image set in the image search result, for example, by combining the quality scores corresponding to the images in the first image set.
The reason why the determined image click feature is used for adjusting the ranking result after the ranking result of the images in the image search result is determined is that the image click feature can directly reflect the image search requirement characteristics of the user, if the ranking position of the images is adjusted through the image click feature too early, most of the images displayed for the user accord with the image search requirement characteristics of the user, so that the images in the image search result are too single, the user is not easy to contact with other types of images, and the user is difficult to provide diversified search experience.
After the ranking result of the images in the image search result is determined, the determined image click feature is used for adjusting the ranking result, so that the image display positions meeting the image search requirement characteristics of the user in the diversified image search results are relatively forward, other types of images with high aesthetic scores or quality scores can be displayed to the user, the image which meets the image search requirement characteristics of the user is preferentially displayed on the premise of keeping the diversity of the image search results, and the image search experience of the user is further improved.
Based on the method for ranking the image search results provided by the embodiment corresponding to fig. 1, the embodiment of the present application provides an apparatus for ranking the image search results, referring to fig. 2, the apparatus includes an obtaining unit 201, an identifying unit 202, a determining unit 203, and a ranking unit 204:
the acquiring unit 201 is configured to acquire an image search result corresponding to a keyword, where the image search result includes multiple images;
the identifying unit 202 is configured to identify a first image set from the image search result according to an image classification model; images in the first set of images belong to a first image category;
the determining unit 203 is configured to determine, through an aesthetic evaluation model corresponding to the first image category, an aesthetic score corresponding to each of the images in the first image set;
the sorting unit 204 is configured to, when sorting the images in the image search result, determine, in combination with the aesthetic scores corresponding to the respective images in the first image set, a sorting position of the images in the first image set in the image search result.
Optionally, the identifying unit is further configured to identify the first image set and the second image set from the image search result according to the image classification model; images in the second set of images belong to a second image category different from the first image category;
the determining unit is further used for determining the aesthetic scores corresponding to the images in the second image set respectively through the aesthetic evaluation models corresponding to the second image categories;
the sorting unit is further used for determining sorting positions of the images in the second image set in the image search results by combining the aesthetic scores corresponding to the images in the second image set when the images in the image search results are sorted.
Optionally, the first image category has a correlation with a type to which the keyword belongs.
Optionally, if the first image type is a person type, the apparatus further includes a face analysis score determining unit and a quality score determining unit:
the face analysis score determining unit is used for determining face analysis scores corresponding to the images in the first image set according to a face analysis model;
the quality score determining unit is used for determining quality scores corresponding to the images in the first image set according to the face analysis scores corresponding to the images in the first image set and the aesthetic scores corresponding to the images in the first image set;
the sorting unit is further configured to determine, in combination with the quality scores corresponding to the images in the first image set, sorting positions of the images in the first image set in the image search result.
Optionally, the image search result is a plurality of images which are obtained by searching according to the keyword and the correlation between the images and the keyword meets a preset condition.
Optionally, if the keyword is input by the target user, the apparatus further includes an image click feature determining unit, an adjusting unit, and a displaying unit:
the image click feature determining unit is used for determining the image click feature of the target user according to the image search behavior of the target user;
the adjusting unit is used for adjusting the sorting result according to the image clicking feature after the sorting result of the images in the image searching result is determined so as to advance the sorting position of the images which accord with the image clicking feature in the image searching result;
and the display unit is used for displaying the image search result according to the adjusted sorting result.
Therefore, when an image search result which corresponds to a keyword and comprises a plurality of images is obtained, a first image set which belongs to a first image category is identified from the images of the image search result according to an image classification model, and the aesthetic scores which correspond to the images in the first image set are determined through an aesthetic evaluation model which corresponds to the first image category. When images in the image search result are sorted, the sorted positions of the images in the first image set in the image search result are determined by combining the aesthetic scores corresponding to the images in the first image set. For the images belonging to an image category, the aesthetic scores of the category can reflect the aesthetic feelings of the images to the user under the category, the possibility that the user selects the images with better aesthetic feelings is higher, therefore, the ranking positions of the images with higher aesthetic scores are advanced, the ranking positions of the images with lower aesthetic scores are advanced, the images with higher aesthetic scores can be preferentially displayed to the user, the probability that the images with the higher aesthetic scores meet the image searching requirements of the user is higher, and the image searching experience of the user is improved.
Based on the foregoing provided quality determination method and apparatus for a traffic channel, the present embodiment provides a quality determination device for a traffic channel, where the quality determination device for a traffic channel may be a terminal device, and fig. 3 is a block diagram of a terminal device 300 according to an exemplary embodiment. For example, the terminal device 300 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, and the like.
Referring to fig. 3, the terminal device 300 may include one or more of the following components: processing component 302, memory 304, power component 306, multimedia component 308, audio component 310, input/output (I/O) interface 312, sensor component 314, and communication component 316.
The processing component 302 generally controls the overall operation of the terminal device 300, such as operations associated with display, telephone calls, data communications, camera operations, and recording operations. The processing elements 302 may include one or more processors 320 to execute instructions to perform all or a portion of the steps of the methods described above. Further, the processing component 302 can include one or more modules that facilitate interaction between the processing component 302 and other components. For example, the processing component 302 can include a multimedia module to facilitate interaction between the multimedia component 308 and the processing component 302.
The memory 304 is configured to store various types of data to support operations at the terminal device 300. Examples of such data include instructions for any application or method operating on device 300, contact data, phonebook data, messages, pictures, videos, and so forth. The memory 304 may be implemented by any type or combination of volatile or non-volatile memory devices, such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks.
The power supply component 306 provides power to the various components of the terminal device 300. The power components 306 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the apparatus 300.
The multimedia component 308 comprises a screen providing an output interface between said terminal device 300 and a user. In some embodiments, the screen may include a Liquid Crystal Display (LCD) and a Touch Panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive an input signal from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 308 includes a front facing camera and/or a rear facing camera. The front camera and/or the rear camera may receive external multimedia data when the terminal device 300 is in an operation mode, such as a photographing mode or a video mode. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.
The audio component 310 is configured to output and/or input audio signals. For example, audio component 310 includes a Microphone (MIC) configured to receive external audio signals when apparatus 300 is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals may further be stored in the memory 304 or transmitted via the communication component 316. In some embodiments, audio component 310 also includes a speaker for outputting audio signals.
The I/O interface 312 provides an interface between the processing component 302 and peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to: a home button, a volume button, a start button, and a lock button.
The communication component 316 is configured to facilitate communication between the terminal device 300 and other devices in a wired or wireless manner. The terminal device 300 may access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication section 316 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 316 further includes a Near Field Communication (NFC) module to facilitate short-range communications. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, infrared data association (IrDA) technology, Ultra Wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.
In an exemplary embodiment, the terminal device 300 may be implemented by one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), controllers, micro-controllers, microprocessors or other electronic components for performing the above-described methods.
In an exemplary embodiment, a non-transitory computer readable storage medium comprising instructions, such as the memory 304 comprising instructions, executable by the processor 320 of the terminal device 300 to perform the above-described method is also provided. For example, the non-transitory computer readable storage medium may be a ROM, a Random Access Memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.
Fig. 4 is a schematic structural diagram of a server in an embodiment of the present invention. The server 400 may vary significantly due to configuration or performance, and may include one or more Central Processing Units (CPUs) 422 (e.g., one or more processors) and memory 432, one or more storage media 430 (e.g., one or more mass storage devices) storing applications 442 or data 444. Wherein the memory 432 and storage medium 430 may be transient or persistent storage. The program stored on the storage medium 430 may include one or more modules (not shown), each of which may include a series of instruction operations for the server. Still further, the central processor 422 may be arranged to communicate with the storage medium 430, and execute a series of instruction operations in the storage medium 430 on the server 400.
The server 400 may also include one or more power supplies 426, one or more wired or wireless network interfaces 450, one or more input-output interfaces 458, one or more keyboards 456, and/or one or more operating systems 441, such as Windows Server, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, and so forth.
A non-transitory computer readable storage medium having instructions therein, which when executed by a processor of a mobile terminal, enable the mobile terminal to perform a method of quality determination of a traffic channel, the method comprising:
acquiring an image search result corresponding to the keyword, wherein the image search result comprises a plurality of images;
identifying a first set of images from the image search results according to an image classification model; images in the first set of images belong to a first image category;
determining an aesthetic score corresponding to each image in the first image set through an aesthetic evaluation model corresponding to the first image category;
when images in the image search result are sorted, the sorted positions of the images in the first image set in the image search result are determined by combining the aesthetic scores corresponding to the images in the first image set.
Those of ordinary skill in the art will understand that: all or part of the steps for realizing the method embodiments can be completed by hardware related to program instructions, the program can be stored in a computer readable storage medium, and the program executes the steps comprising the method embodiments when executed; and the aforementioned storage medium may be at least one of the following media: various media that can store program codes, such as read-only memory (ROM), RAM, magnetic disk, or optical disk.
It should be noted that, in the present specification, all the embodiments are described in a progressive manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the apparatus and system embodiments, since they are substantially similar to the method embodiments, they are described in a relatively simple manner, and reference may be made to some of the descriptions of the method embodiments for related points. The above-described embodiments of the apparatus and system are merely illustrative, and the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
The above description is only one specific embodiment of the present application, but the scope of the present application is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present application should be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A method for ranking image search results, the method comprising:
acquiring an image search result corresponding to the keyword, wherein the image search result comprises a plurality of images;
identifying a first set of images from the image search results according to an image classification model; images in the first set of images belong to a first image category;
determining an aesthetic score corresponding to each image in the first image set through an aesthetic evaluation model corresponding to the first image category;
when images in the image search result are sorted, the sorted positions of the images in the first image set in the image search result are determined by combining the aesthetic scores corresponding to the images in the first image set.
2. The method of claim 1, wherein identifying a first set of images from the image search results according to an image classification model comprises:
identifying the first image set and the second image set from the image search results according to the image classification model; images in the second set of images belong to a second image category different from the first image category;
the method further comprises the following steps:
determining an aesthetic rating score corresponding to each image in the second image set through an aesthetic evaluation model corresponding to the second image category;
and when the images in the image search result are sorted, determining the sorting positions of the images in the second image set in the image search result by combining the aesthetic scores corresponding to the images in the second image set respectively.
3. The method of claim 1, wherein the first image category has a correlation with a type to which the keyword belongs.
4. The method of claim 1, wherein if the first image category is a human category, the method further comprises:
determining face analysis scores corresponding to the images in the first image set according to a face analysis model;
determining quality scores corresponding to the images in the first image set according to the face analysis scores corresponding to the images in the first image set and the aesthetic scores corresponding to the images in the first image set;
the determining the ranking position of the images in the first image set in the image search result by combining the aesthetic scores corresponding to the images in the first image set respectively comprises:
and determining the sorting position of the images in the first image set in the image search result by combining the quality scores corresponding to the images in the first image set respectively.
5. The method according to claim 1, wherein the image search result is a plurality of images which are obtained by searching according to the keyword and have a correlation with the keyword satisfying a preset condition.
6. The method of any one of claims 1-5, wherein if the keyword is input by a target user, the method further comprises:
determining the image click characteristics of the target user according to the image search behavior of the target user;
after the sorting result of the images in the image searching result is determined, adjusting the sorting result according to the image clicking feature so as to advance the sorting position of the images which accord with the image clicking feature in the image searching result;
and displaying the image search result according to the adjusted sorting result.
7. An apparatus for ranking image search results, the apparatus comprising an acquisition unit, a recognition unit, a determination unit, and a ranking unit:
the acquisition unit is used for acquiring an image search result corresponding to the keyword, and the image search result comprises a plurality of images;
the identification unit is used for identifying a first image set from the image search results according to an image classification model; images in the first set of images belong to a first image category;
the determining unit is used for determining the aesthetic scores corresponding to the images in the first image set respectively through the aesthetic evaluation models corresponding to the first image categories;
and the sorting unit is used for determining the sorting positions of the images in the first image set in the image search results by combining the aesthetic scores corresponding to the images in the first image set when the images in the image search results are sorted.
8. The apparatus according to claim 7, wherein the identifying unit is further configured to identify the first set of images and the second set of images from the image search result according to the image classification model; images in the second set of images belong to a second image category different from the first image category;
the determining unit is further used for determining the aesthetic scores corresponding to the images in the second image set respectively through the aesthetic evaluation models corresponding to the second image categories;
the sorting unit is further used for determining sorting positions of the images in the second image set in the image search results by combining the aesthetic scores corresponding to the images in the second image set when the images in the image search results are sorted.
9. An apparatus for ranking image search results, comprising a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured for execution by one or more processors the one or more programs including instructions for:
acquiring an image search result corresponding to the keyword, wherein the image search result comprises a plurality of images;
identifying a first set of images from the image search results according to an image classification model; images in the first set of images belong to a first image category;
determining an aesthetic score corresponding to each image in the first image set through an aesthetic evaluation model corresponding to the first image category;
when images in the image search result are sorted, the sorted positions of the images in the first image set in the image search result are determined by combining the aesthetic scores corresponding to the images in the first image set.
10. A machine-readable medium having stored thereon instructions, which when executed by one or more processors, cause an apparatus to perform a method of ranking image search results as recited in one or more of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811612143.2A CN111382295B (en) | 2018-12-27 | 2018-12-27 | Image search result ordering method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811612143.2A CN111382295B (en) | 2018-12-27 | 2018-12-27 | Image search result ordering method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111382295A true CN111382295A (en) | 2020-07-07 |
| CN111382295B CN111382295B (en) | 2024-04-30 |
Family
ID=71219559
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811612143.2A Active CN111382295B (en) | 2018-12-27 | 2018-12-27 | Image search result ordering method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111382295B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114328359A (en) * | 2020-09-27 | 2022-04-12 | 广州市久邦数码科技有限公司 | Playing method and system of audio electronic book |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102332034A (en) * | 2011-10-21 | 2012-01-25 | 中国科学院计算技术研究所 | Portrait picture retrieval method and device |
| CN102799635A (en) * | 2012-06-27 | 2012-11-28 | 天津大学 | Image set ordering method driven by user |
| CN103699612A (en) * | 2013-12-13 | 2014-04-02 | 中国科学院深圳先进技术研究院 | Image retrieval ranking method and device |
| CN103988202A (en) * | 2011-11-25 | 2014-08-13 | 微软公司 | Image attractiveness based indexing and searching |
| US20140351264A1 (en) * | 2013-05-21 | 2014-11-27 | Xerox Corporation | Methods and systems for ranking images using semantic and aesthetic models |
| US20160098844A1 (en) * | 2014-10-03 | 2016-04-07 | EyeEm Mobile GmbH | Systems, methods, and computer program products for searching and sorting images by aesthetic quality |
| CN107153838A (en) * | 2017-04-19 | 2017-09-12 | 中国电子科技集团公司电子科学研究院 | A kind of photo automatic grading method and device |
| CN107729384A (en) * | 2017-09-18 | 2018-02-23 | 维沃移动通信有限公司 | Image ranking method and mobile terminal |
| EP3321855A1 (en) * | 2016-11-15 | 2018-05-16 | Houzz, Inc. | Aesthetic search engine |
| CN108288027A (en) * | 2017-12-28 | 2018-07-17 | 新智数字科技有限公司 | A kind of detection method of picture quality, device and equipment |
| WO2018192245A1 (en) * | 2017-04-19 | 2018-10-25 | 中国电子科技集团公司电子科学研究院 | Automatic scoring method for photo based on aesthetic assessment |
| CN108897685A (en) * | 2018-06-28 | 2018-11-27 | 百度在线网络技术（北京）有限公司 | Method for evaluating quality, device, server and the medium of search result |
-
2018
- 2018-12-27 CN CN201811612143.2A patent/CN111382295B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102332034A (en) * | 2011-10-21 | 2012-01-25 | 中国科学院计算技术研究所 | Portrait picture retrieval method and device |
| CN103988202A (en) * | 2011-11-25 | 2014-08-13 | 微软公司 | Image attractiveness based indexing and searching |
| CN102799635A (en) * | 2012-06-27 | 2012-11-28 | 天津大学 | Image set ordering method driven by user |
| US20140351264A1 (en) * | 2013-05-21 | 2014-11-27 | Xerox Corporation | Methods and systems for ranking images using semantic and aesthetic models |
| CN103699612A (en) * | 2013-12-13 | 2014-04-02 | 中国科学院深圳先进技术研究院 | Image retrieval ranking method and device |
| US20160098844A1 (en) * | 2014-10-03 | 2016-04-07 | EyeEm Mobile GmbH | Systems, methods, and computer program products for searching and sorting images by aesthetic quality |
| EP3321855A1 (en) * | 2016-11-15 | 2018-05-16 | Houzz, Inc. | Aesthetic search engine |
| CN107153838A (en) * | 2017-04-19 | 2017-09-12 | 中国电子科技集团公司电子科学研究院 | A kind of photo automatic grading method and device |
| WO2018192245A1 (en) * | 2017-04-19 | 2018-10-25 | 中国电子科技集团公司电子科学研究院 | Automatic scoring method for photo based on aesthetic assessment |
| CN107729384A (en) * | 2017-09-18 | 2018-02-23 | 维沃移动通信有限公司 | Image ranking method and mobile terminal |
| CN108288027A (en) * | 2017-12-28 | 2018-07-17 | 新智数字科技有限公司 | A kind of detection method of picture quality, device and equipment |
| CN108897685A (en) * | 2018-06-28 | 2018-11-27 | 百度在线网络技术（北京）有限公司 | Method for evaluating quality, device, server and the medium of search result |
Non-Patent Citations (5)
| Title |
|---|
| CHAORAN CUI;HUIDI FANG;XIANG DENG;XIUSHAN NIE;HONGSHUAI DAI;YILONG YIN: "Distribution-oriented Aesthetics Assessment for Image Search", SIGIR\'17: PROCEEDINGS OF THE 40TH INTERNATIONAL ACM SIGIR CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION RETRIEVAL, pages 1013 - 1016 * |
| MITHUN BISWAS; RAFIQUL ISLAM; GAUTAM KUMAR SHOM; NABEEL MOHAMMED; SIFAT MOMEN; NAFEES MANSOOR; ANOWARUL ABEDIN: "Application of image retrieval for aesthetic evaluation and improvement suggestion of isolated Bangla handwritten characters", 2017 IEEE INTERNATIONAL CONFERENCE ON SIGNAL AND IMAGE PROCESSING APPLICATIONS (ICSIPA), 30 November 2017 (2017-11-30), pages 149 - 154 * |
| 杨文雅;宋广乐;崔超然;尹义龙;: "基于语义感知的图像美学质量评估方法", 计算机应用, vol. 38, no. 11, pages 3216 - 3220 * |
| 王伟凝;蚁静缄;贺前华: "可计算图像美学研究进展", 中国图象图形学报, vol. 17, no. 08, 16 August 2012 (2012-08-16), pages 893 - 901 * |
| 赵小艳;刘宏哲;袁家政;杨少鹏;: "图像重排序技术的研究进展", 计算机科学, no. 05, 15 May 2018 (2018-05-15), pages 15 - 23 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114328359A (en) * | 2020-09-27 | 2022-04-12 | 广州市久邦数码科技有限公司 | Playing method and system of audio electronic book |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111382295B (en) | 2024-04-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11622141B2 (en) | Method and apparatus for recommending live streaming room | |
| US11520824B2 (en) | Method for displaying information, electronic device and system | |
| US20170161582A1 (en) | Image processing method, device and medium | |
| CN107315487B (en) | Input processing method and device and electronic equipment | |
| CN107784045B (en) | Quick reply method and device for quick reply | |
| CN109918565B (en) | Processing method and device for search data and electronic equipment | |
| CN107229403B (en) | Information content selection method and device | |
| CN107291772B (en) | Search access method and device and electronic equipment | |
| CN111556352B (en) | Multimedia resource sharing method and device, electronic equipment and storage medium | |
| CN111970566A (en) | Video playing method and device, electronic equipment and storage medium | |
| CN106815291B (en) | Search result item display method and device and search result item display device | |
| CN111382339A (en) | Search processing method and device and search processing device | |
| US11546663B2 (en) | Video recommendation method and apparatus | |
| CN107728877B (en) | Application recommendation method and mobile terminal | |
| CN113407828A (en) | Searching method, searching device and searching device | |
| CN107491453B (en) | Method and device for identifying cheating web pages | |
| CN112015277B (en) | Information display method and device and electronic equipment | |
| CN111382295B (en) | Image search result ordering method and device | |
| CN110046308B (en) | Sequencing strategy determination method and device and electronic equipment | |
| CN112115341A (en) | Content display method, device, terminal, server, system and storage medium | |
| CN111752436A (en) | Recommendation method and device and recommendation device | |
| CN108205534B (en) | Skin resource display method and device and electronic equipment | |
| CN113965792A (en) | Video display method and device, electronic equipment and readable storage medium | |
| CN107301188B (en) | Method for acquiring user interest and electronic equipment | |
| CN112784858B (en) | Image data processing method and 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |