CN112395447A - Label system updating method, device and equipment based on picture and storage medium - Google Patents
Label system updating method, device and equipment based on picture and storage medium Download PDFInfo
- Publication number
- CN112395447A CN112395447A CN202011360402.4A CN202011360402A CN112395447A CN 112395447 A CN112395447 A CN 112395447A CN 202011360402 A CN202011360402 A CN 202011360402A CN 112395447 A CN112395447 A CN 112395447A
- Authority
- CN
- China
- Prior art keywords
- picture
- initial
- pictures
- initial picture
- recalled
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 59
- 238000004590 computer program Methods 0.000 claims description 11
- 238000002372 labelling Methods 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 10
- 230000008569 process Effects 0.000 description 9
- 230000006870 function Effects 0.000 description 8
- 238000004891 communication Methods 0.000 description 6
- 230000003287 optical effect Effects 0.000 description 6
- 238000012545 processing Methods 0.000 description 6
- 241000227653 Lycopersicon Species 0.000 description 4
- 235000007688 Lycopersicon esculentum Nutrition 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 239000013307 optical fiber Substances 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/5866—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using information manually generated, e.g. tags, keywords, comments, manually generated location and time information
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Library & Information Science (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the disclosure discloses a method, a device, equipment and a storage medium for updating a label system based on pictures. The method comprises the following steps: acquiring an initial label system to be updated, wherein the initial label system comprises at least two labels and an initial picture set corresponding to each label; determining a first recall picture of a first initial picture in a first initial picture set and a second recall picture of a second initial picture in a second initial picture set; determining similarity scores of the two initial picture sets according to the first recall picture and the second recall picture; and if the similarity score meets a first preset condition, combining the two initial image sets and the corresponding labels to obtain a target label system. According to the scheme, the recall picture can be automatically determined according to the initial picture, the two initial picture sets and the corresponding labels are combined according to the recall picture, the automatic updating of the initial label system is achieved, manual operation is not needed, the labor cost and the time cost are saved, and the updating efficiency is improved.
Description
Technical Field
The embodiment of the disclosure relates to the technical field of computers, in particular to a method, a device, equipment and a storage medium for updating a label system based on pictures.
Background
The label system is a label set obtained by analyzing data, screening out multiple types of key information contained in the data, and allocating corresponding labels to each type of key information. With the development of deep learning, more and more fields need fine-grained label system support. For example, in a picture search scene, a user may search for a tag system corresponding to a picture as needed to obtain related information. As the exponential growth of video and picture data, the label system needs to be updated to meet the application requirements.
At the present stage, the label system is mainly updated in a manual labeling mode, a large amount of labor and time are needed, and the efficiency is low.
BRIEF SUMMARY OF THE PRESENT DISCLOSURE
The embodiment of the disclosure provides a method, a device, equipment and a storage medium for updating a label system based on pictures, which can optimize the existing label system updating scheme.
In a first aspect, an embodiment of the present disclosure provides a method for updating a label system based on an image, including:
acquiring an initial label system to be updated, wherein the initial label system comprises at least two labels and an initial picture set corresponding to each label, and the initial picture set comprises at least one initial picture;
determining a first recall picture of a first initial picture in the first initial picture set and a second recall picture of a second initial picture in the second initial picture set aiming at the first initial picture set and the second initial picture set; the first initial picture set and the second initial picture set are any two picture sets in the initial picture sets, and the first recalled picture and the second recalled picture are pictures which are similar to the first initial picture and the second initial picture respectively in a picture library;
determining similarity scores of the first initial picture set and the second initial picture set according to the first recall picture and the second recall picture;
and if the similarity score meets a first preset condition, merging the first initial image set and the second initial image set and merging a first label corresponding to the first initial image set and a second label corresponding to the second initial image set to obtain a target label system, wherein the target label system is used as an updating result of the initial label system.
In a second aspect, an embodiment of the present disclosure further provides a device for updating a label system based on a picture, including:
the system comprises an acquisition module, a updating module and a display module, wherein the acquisition module is used for acquiring an initial label system to be updated, the initial label system comprises at least two labels and an initial picture set corresponding to each label, and the initial picture set comprises at least one initial picture;
a recalled picture determining module, configured to determine, for a first initial picture set and a second initial picture set, a first recalled picture of a first initial picture in the first initial picture set and a second recalled picture of a second initial picture in the second initial picture set; the first initial picture set and the second initial picture set are any two picture sets in the initial picture sets, and the first recalled picture and the second recalled picture are pictures which are similar to the first initial picture and the second initial picture respectively in a picture library;
a similarity score determining module, configured to determine similarity scores of the first initial picture set and the second initial picture set according to the first recalled picture and the second recalled picture;
and the updating module is used for merging the first initial image set and the second initial image set and merging a first label corresponding to the first initial image set and a second label corresponding to the second initial image set to obtain a target label system as an updating result of the initial label system if the similarity score meets a first preset condition.
In a third aspect, an embodiment of the present disclosure further provides an electronic device, including:
one or more processors;
a memory for storing one or more programs;
the one or more programs, when executed by the one or more processors, implement the picture-based tag hierarchy updating method as described in the first aspect.
In a fourth aspect, an embodiment of the present disclosure further provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the method for updating a label system based on pictures according to the first aspect.
The embodiment of the disclosure provides a method, a device, equipment and a storage medium for updating a label system based on pictures, wherein an initial label system to be updated is obtained, the initial label system comprises at least two labels and an initial picture set corresponding to each label, and the initial picture set comprises at least one initial picture; determining a first recall picture of a first initial picture in the first initial picture set and a second recall picture of a second initial picture in the second initial picture set aiming at the first initial picture set and the second initial picture set; determining similarity scores of the first initial picture set and the second initial picture set according to the first recall picture and the second recall picture; and if the similarity score meets a first preset condition, merging the first initial image set and the second initial image set and merging a first label corresponding to the first initial image set and a second label corresponding to the second initial image set to obtain a target label system, wherein the target label system is used as an updating result of the initial label system. According to the scheme, the recall picture corresponding to the initial picture can be automatically determined according to the initial picture, the similarity scores of the two initial picture sets are determined according to the recall picture, when the similarity scores meet the first preset condition, the two initial picture sets and the corresponding labels are combined, the automatic updating of the initial label system is achieved, manual operation is not needed, the labor cost and the time cost are saved, and the updating efficiency is improved.
Drawings
The above and other features, advantages and aspects of various embodiments of the present disclosure will become more apparent by referring to the following detailed description when taken in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numbers refer to the same or similar elements. It should be understood that the drawings are schematic and that elements and features are not necessarily drawn to scale.
Fig. 1 is a flowchart of a method for updating a tag hierarchy based on a picture according to a first embodiment of the present disclosure;
fig. 2 is a flowchart of a method for updating a label system based on a picture according to a second embodiment of the present disclosure;
fig. 3 is a flowchart of a method for updating a label hierarchy based on a picture according to a third embodiment of the present disclosure;
fig. 4 is a schematic diagram of an implementation process of a tag system updating method based on a picture according to a third embodiment of the present disclosure;
fig. 5 is a structural diagram of a tag system updating apparatus based on pictures according to a fourth embodiment of the disclosure;
fig. 6 is a structural diagram of an electronic device according to a fifth embodiment of the present disclosure.
Detailed Description
Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While certain embodiments of the present disclosure are shown in the drawings, it is to be understood that the present disclosure may be embodied in various forms and should not be construed as limited to the embodiments set forth herein, but rather are provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the disclosure are for illustration purposes only and are not intended to limit the scope of the disclosure.
It should be understood that the various steps recited in the method embodiments of the present disclosure may be performed in a different order, and/or performed in parallel. Moreover, method embodiments may include additional steps and/or omit performing the illustrated steps. The scope of the present disclosure is not limited in this respect.
The term "include" and variations thereof as used herein are open-ended, i.e., "including but not limited to". The term "based on" is "based, at least in part, on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Relevant definitions for other terms will be given in the following description.
It should be noted that the terms "first", "second", and the like in this disclosure are only used for distinguishing different objects, and are not used for limiting the order or interdependence relationship of the objects.
It is noted that references to "a", "an", and "the" modifications in this disclosure are intended to be illustrative rather than limiting, and that those skilled in the art will recognize that "one or more" may be used unless the context clearly dictates otherwise.
The names of messages or information exchanged between devices in the embodiments of the present disclosure are for illustrative purposes only, and are not intended to limit the scope of the messages or information.
Example one
Fig. 1 is a flowchart of a method for updating a tag system based on a picture according to an embodiment of the present disclosure, where this embodiment is applicable to updating a tag system obtained based on a picture, and the method may be executed by a tag system updating apparatus based on a picture, where the apparatus may be implemented in a software and/or hardware manner, and the apparatus may be configured in an electronic device, and the electronic device may be a terminal with a data processing function, such as a mobile terminal, e.g., a mobile phone, a tablet, a notebook, and the like, or a fixed terminal, e.g., a desktop, or a server. As shown in fig. 1, the method specifically includes the following steps:
and S110, acquiring an initial label system to be updated.
The initial label system comprises at least two labels and an initial picture set corresponding to each label, and the initial picture set comprises at least one initial picture. The initial label system of this embodiment is obtained by performing cluster analysis on the pictures, and optionally, the original pictures may be obtained, the features of the original pictures are extracted, the original pictures are clustered according to the features, an initial picture set and labels corresponding to the initial picture set are obtained, and the initial label system may be formed according to the initial label set and the labels corresponding to the initial label set. The original picture can be obtained from the internet in a crawler mode, and can also be obtained based on a picture search engine. The features of the original picture may include some key features of the original picture, such as color features, texture features, contour features, and the like of the original picture. And performing clustering analysis on the original pictures according to the characteristics to obtain an initial picture set so as to store different clustering results. And associating the tags with the corresponding initial image sets to obtain an initial tag system.
For example, the initial tag hierarchy may be denoted as D ═ S1,S2,…,Sn},Sk={p1,p2,…,piIn which D represents the initial label system, SkIndicates the initial picture set corresponding to the kth label, where k is 1,2, …, n, n is the number of labels included in the initial label system, i is the number of initial picture sets, i is the initial picture set SkNumber of initial pictures contained, piRepresenting an initial set of pictures SkThe ith initial picture in (1), which may be the original picture as described above.
It can be understood that, with the development of internet technology, the data volume of pictures is also increasing, and the requirements of users on pictures are also changing, for example, in some scenarios, users want to accurately obtain some fine-grained pictures, and in some scenarios, only coarse-grained pictures need to be obtained. The existing initial label system can only meet the requirement of the user on the coarse granularity and cannot meet the requirement of the user on the fine granularity, so that the existing initial label system needs to be updated, and the updated initial label system can meet the requirement of obtaining the user on the fine-granularity picture.
S120, aiming at a first initial picture set and a second initial picture set, determining a first recall picture of a first initial picture in the first initial picture set and a second recall picture of a second initial picture in the second initial picture set.
The first initial picture set and the second initial picture set are any two picture sets in the initial picture sets, and the first recalled picture and the second recalled picture are pictures similar to the first initial picture and the second initial picture in a picture library respectively. Optionally, when the initial tag system only includes two initial image sets, one of the initial image sets may be used as a first initial image set, and the other initial image set may be used as a second initial image set, and when the number of the initial image sets included in the initial tag system is greater than 2, any two initial image sets may be selected from the initial image sets to be used as the first initial image set and the second initial image set, respectively. The first initial picture is a picture contained in the first initial picture set, and the second initial picture is a picture contained in the second initial picture set. The recall picture may be a picture similar to the initial picture obtained from a picture library, and the embodiment refers to the picture similar to the first initial picture obtained from the picture library as a first recall picture, and refers to the picture similar to the second initial picture obtained from the picture library as a second recall picture. The picture library is used for storing pictures and reference picture characteristics corresponding to the pictures, and each picture and the reference picture characteristics corresponding to the picture can be stored in association, for example, the picture A and the reference picture characteristics of the picture A are stored in association. The reference picture features may be some key features of the picture, such as color features, texture features, contour features, and the like of the picture, and the corresponding picture may be found by referring to the picture features.
The determination process of the first recall picture and the second recall picture is similar, and the first recall picture is taken as an example. Optionally, the picture feature of the first initial picture may be extracted, the picture feature is matched with each reference picture feature in the picture library, the similarity score between the picture feature and each reference picture feature is determined, when the similarity score is greater than a set threshold, the reference picture feature corresponding to the similarity score is used as the picture feature similar to the picture feature, and the picture corresponding to the reference picture feature is used as the first recall picture. The embodiment does not limit the size of the set threshold. Therefore, a corresponding first recall picture can be obtained based on each first initial picture contained in the first initial picture set. Optionally, the first recall picture may be stored in a recall picture set, where the recall picture set is used to store recall pictures, and in order to facilitate distinction, the first recall picture corresponding to each first initial picture may be stored in one recall picture set.
S130, according to the first recall picture and the second recall picture, determining similarity scores of the first initial picture set and the second initial picture set.
The similarity score is used to indicate the similarity between the first initial picture set and the second initial picture set, and may be represented by a value between 0 and 1, where the larger the value, the more similar the first initial picture set and the second initial picture set are. In one example, the first recall picture and the second recall picture may be input into a similarity model, and similarity scores for the first initial picture set and the second initial picture set may be output by the similarity model. The similarity model may adopt a neural network model, and the embodiment does not limit the specific structure of the neural network model. In one example, a number of recall pictures belonging to the other initial picture set may also be determined, based on which similarity scores for the two initial picture sets are determined. For example, the number of the first recalled pictures belonging to the second initial picture set and the number of the second recalled pictures belonging to the first initial picture set may be determined, so as to obtain the total number of the first recalled pictures and the second recalled pictures belonging to the counterpart initial picture set, and the similarity scores of the first initial picture set and the second initial picture set may be determined according to the total number and the numbers of the first initial pictures and the second initial pictures.
And S140, if the similarity score meets a first preset condition, merging the first initial image set and the second initial image set, and merging a first label corresponding to the first initial image set and a second label corresponding to the second initial image set to obtain a target label system, wherein the target label system is used as an updating result of the initial label system.
The first preset condition may be that the similarity score is greater than or equal to a set threshold. Specifically, if the similarity scores of the first initial picture set and the second initial picture set are greater than or equal to the set threshold, it is considered that the similarity between the first initial picture set and the second initial picture set is relatively large, so that the first initial picture set and the second initial picture set can be merged, and the first tag corresponding to the first initial picture set and the second tag corresponding to the second initial picture set can be merged. If the similarity scores of the first initial picture set and the second initial picture set are smaller than the set threshold, the similarity of the first initial picture set and the second initial picture set is considered to be smaller, and therefore the first initial picture set and the second initial picture set can be kept unchanged.
For example, the first tag is a tomato, the second tag is a tomato, and accordingly, the first initial picture set is pictures corresponding to the tomato, the second initial picture set is pictures corresponding to the tomato, and the first initial picture set and the second initial picture set can be merged into a new set by determining that the similarity scores of the first initial picture set and the second initial picture set meet the first preset condition, where the new set includes both the first initial picture in the first initial picture set and the second initial picture in the second initial picture set. The label to which the new set corresponds may be the first label or the second label.
The embodiment of the disclosure provides a tag system updating method based on pictures, which includes obtaining an initial tag system to be updated, where the initial tag system includes at least two tags and an initial picture set corresponding to each tag, and the initial picture set includes at least one initial picture; determining a first recall picture of a first initial picture in the first initial picture set and a second recall picture of a second initial picture in the second initial picture set aiming at the first initial picture set and the second initial picture set; determining similarity scores of the first initial picture set and the second initial picture set according to the first recall picture and the second recall picture; and if the similarity score meets a first preset condition, merging the first initial image set and the second initial image set and merging a first label corresponding to the first initial image set and a second label corresponding to the second initial image set to obtain a target label system, wherein the target label system is used as an updating result of the initial label system. According to the scheme, the recall picture corresponding to the initial picture can be automatically determined according to the initial picture, the similarity scores of the two initial picture sets are determined according to the recall picture, when the similarity scores meet the first preset condition, the two initial picture sets and the corresponding labels are combined, the automatic updating of the initial label system is achieved, manual operation is not needed, the labor cost and the time cost are saved, and the updating efficiency is improved.
Example two
Fig. 2 is a flowchart of a method for updating a label system based on a picture according to a second embodiment of the present disclosure, where the present embodiment is optimized based on the foregoing embodiment, and referring to fig. 2, the method may include the following steps:
and S210, acquiring an initial label system to be updated.
S220, aiming at a first initial picture set and a second initial picture set, determining a first recall picture of a first initial picture in the first initial picture set and a second recall picture of a second initial picture in the second initial picture set.
The embodiment does not limit the determination order of the first recall picture and the second recall picture, for example, the first recall picture corresponding to the first initial picture may be determined first, then the second recall picture corresponding to the second initial picture may be determined, the second recall picture corresponding to the second initial picture may be determined first, then the first recall picture corresponding to the first initial picture may be determined, and the first recall picture corresponding to the first initial picture and the second recall picture corresponding to the second initial picture may be determined at the same time.
In one example, S220 may include:
s2201, extracting first picture features of the first initial picture and second picture features of the second initial picture respectively.
The first picture feature is a key feature of the first initial picture, and may include, for example, a color feature, a texture feature, and a contour feature of the first initial picture. The second picture features are key features of the second initial picture, and may include, for example, color features, texture features, contour features, and the like of the second initial picture. The embodiment does not limit the specific feature extraction method.
S2202, determining first similarity scores of the first picture feature and the reference picture feature and second similarity scores of the second picture feature and the reference picture feature.
The reference picture features are picture features in the picture library, and the picture library is used for storing the corresponding relation between the reference picture features and pictures. For the related content of the photo library, reference may be made to the above embodiments, which are not described herein again. The first similarity score is used for representing the degree of similarity between the first picture feature and each reference picture feature in the picture library, and the second similarity score is used for representing the degree of similarity between the second picture feature and each reference picture feature in the picture library. The first similarity score and the second similarity score can be represented by numerical values between 0 and 1, and the larger the numerical value is, the higher the similarity degree is.
The determination process of the first similarity score and the second similarity score is similar, and the first similarity score is taken as an example. In one example, the first picture feature and the reference picture feature may be input into a similarity model, and the similarity model outputs a first similarity score between the first picture feature and the reference picture feature, and in consideration of the fact that the number of pictures stored in the picture library is large and the number of corresponding reference picture features is large, in order to improve efficiency, the first picture feature and the batch of reference picture features may be input into the similarity model, and the similarity model outputs the first similarity score between the first picture feature and the batch of reference picture features. In one example, euclidean distances between the first picture feature and the reference picture features may also be determined separately, and the first similarity score between the first picture feature and the reference picture features may be determined based on the euclidean distances.
S2203, if the first similarity score is larger than or equal to a first set threshold, taking a picture corresponding to the reference picture characteristic as a first recalled picture; and if the second similarity score is larger than or equal to a second set threshold value, taking the picture corresponding to the reference picture characteristic as a second recalled picture.
The embodiment does not limit the size of the first set threshold and the second set threshold, and the first set threshold and the second set threshold may be the same or different in actual application. Optionally, when the first similarity score is greater than or equal to the first set threshold, the reference picture feature corresponding to the first similarity score may be used as a picture feature similar to the first picture feature, and a picture corresponding to the reference picture feature may be used as a first recalled picture and stored in the corresponding recalled picture set. Similarly, when the second similarity score is greater than or equal to the second set threshold, the reference picture feature corresponding to the second similarity score may be regarded as a picture feature similar to the second picture feature, and a picture corresponding to the reference picture feature may be regarded as a second recall picture and stored in the corresponding recall picture set.
S230, determining a first number of the first recalled pictures that belong to the second initial picture set and a second number of the second recalled pictures that belong to the first initial picture set.
In this embodiment, the similarity scores of two initial picture sets are determined based on the number of initial picture sets belonging to the other initial picture set in the recalled pictures. Specifically, for each first initial picture, the number of the first recalled pictures corresponding to the first initial picture, which belong to the second initial picture set, may be determined, so as to obtain the total number of the first recalled pictures corresponding to each first initial picture, which belongs to the second initial picture set. Similarly, for each second initial picture, the number of the second recalled pictures corresponding to the second initial picture that belong to the first initial picture set may be determined, so as to obtain the total number of the second recalled pictures corresponding to each second initial picture that belong to the first initial picture set, where this total number is referred to as a second number in an embodiment.
S240, determining similarity scores of the first initial picture set and the second initial picture set according to the first quantity and the second quantity, and a first initial quantity of first initial pictures contained in the first initial picture set and a second initial quantity of second initial pictures contained in the second initial picture set.
The first initial number is the number of first initial pictures contained in the first initial picture set before the initial label system is updated, and the second initial number is the number of second initial pictures contained in the second initial picture set before the initial label system is updated. Optionally, the first number and the second number may be accumulated to obtain a third number, and the first initial number and the second initial number may be accumulated to obtain a third initial number; and taking the ratio of the third number to the third initial number as the similarity score of the first initial picture set and the second initial picture set. For example, the first initial picture set includes m first initial pictures, and the second initial picture set includes o second initial pictures, where w recalling pictures belong to each other in total, that is, the sum of the total number of the first recalling pictures corresponding to the m first initial pictures and the total number of the second recalling pictures corresponding to the o second initial pictures, which belong to the second initial picture set, is w, then the similarity scores of the first initial picture set and the second initial picture set may be represented as: w/(m + o), w is m + o at the maximum.
And S250, judging whether the similarity score meets a first preset condition, if so, executing S260, and otherwise, executing S270.
Taking the first preset condition as an example that the similarity score is greater than or equal to the set threshold, when the similarity score between the first initial picture set and the second initial picture set is greater than or equal to the set threshold, indicating that the similarity degree between the first initial picture set and the second initial picture set is higher, S260 may be executed to merge the first initial picture set and the second initial picture set and the tags corresponding to the two sets. Otherwise, data cleaning can be carried out on the first initial picture set and the second initial picture set, and noise in the first initial picture set and the second initial picture set is removed. The removal process can be seen in the following steps.
S260, merging the first initial picture set and the second initial picture set, and merging a first label corresponding to the first initial picture set and a second label corresponding to the second initial picture set.
S270, determining a first picture quantity belonging to the first initial picture set in a first recall picture corresponding to the first initial picture and a second picture quantity belonging to the second initial picture set in a second recall picture corresponding to the second initial picture.
Optionally, whether the corresponding initial picture is noise may be determined according to the number of the initial picture sets in which the corresponding initial pictures are located in the recalled pictures, and when it is determined that the initial picture is noise, the initial picture may be deleted from the initial picture set in which the initial picture is located, so as to optimize the corresponding initial picture set. Specifically, a first number of pictures belonging to the first initial picture set in the first recalled picture and a second number of pictures belonging to the second initial picture set in the second recalled picture may be respectively determined, and then whether the first initial picture is noise or not may be determined according to the first number of pictures, and whether the second initial picture is noise or not may be determined according to the second number of pictures.
S280, if the number of the first pictures is smaller than a third set threshold value, deleting the first initial pictures from the first initial picture set; and if the number of the second pictures is smaller than a fourth set threshold value, deleting the second initial pictures from the second initial picture set.
Specifically, if the number of first pictures is less than the third set threshold, the first initial picture may be considered as noise, and the first initial picture may be deleted from the first initial picture set, and similarly, if the number of second pictures is less than the fourth set threshold, the second initial picture may be considered as noise, and the second initial picture may be deleted from the second initial picture set, so that the first initial picture set and the second initial picture set may be optimized. The magnitudes of the third set threshold and the fourth set threshold may be set according to actual conditions.
The updating process can be repeatedly executed to continuously optimize the initial label system and meet the requirements of users.
The second embodiment of the present disclosure provides a method for updating a tag system based on a picture, which may determine, based on the above embodiments, a recall picture corresponding to an initial picture from a picture library according to the initial picture in the initial picture set, determine similarity scores of two initial picture sets according to the recall picture, and further implement merging of the initial picture sets and merging of tags according to the similarity scores; on the basis, when the two initial picture sets cannot be combined, whether the corresponding initial picture is noise or not can be further determined according to the number of pictures belonging to the label in the recalled pictures, and when the initial picture is noise, the initial picture is deleted from the initial picture set, so that the corresponding initial picture set is optimized, manual operation is not needed, the labor cost and the time cost are reduced, and the updating efficiency is improved.
EXAMPLE III
Fig. 3 is a flowchart of a method for updating a tag system based on pictures according to a third embodiment of the present disclosure, where this embodiment is optimized based on the foregoing embodiment, and fig. 3 exemplarily shows another way of optimizing a first initial picture set and a second initial picture set, where S310 to S340 are the same as S210 to S240 in the foregoing embodiment, and are not described again here. Referring to fig. 3, the method may include the steps of:
s310, obtaining an initial label system to be updated.
S320, aiming at a first initial picture set and a second initial picture set, determining a first recall picture of a first initial picture in the first initial picture set and a second recall picture of a second initial picture in the second initial picture set.
S330, determining a first number of the first recalled pictures belonging to the second initial picture set and a second number of the second recalled pictures belonging to the first initial picture set.
S340, determining similarity scores of the first initial picture set and the second initial picture set according to the first quantity and the second quantity, and a first initial quantity of first initial pictures included in the first initial picture set and a second initial quantity of second initial pictures included in the second initial picture set.
And S350, judging whether the similarity score meets a first preset condition, if so, executing S360, and otherwise, executing S370.
S360, merging the first initial picture set and the second initial picture set, and merging a first label corresponding to the first initial picture set and a second label corresponding to the second initial picture set to obtain a target label system as an updating result of the initial label system.
S370, determining a first ratio of the number of pictures belonging to the first initial picture set in a first recall picture corresponding to the first initial picture to the number of pictures belonging to the first recall picture, and a second ratio of the number of pictures belonging to the second initial picture set in a second recall picture corresponding to the second initial picture to the number of pictures belonging to the second recall picture.
In an example, it may also be determined whether a certain initial picture is noise according to a ratio of the number of initial picture sets in which the initial picture belongs to the recalled pictures corresponding to the initial picture to the number of recalled pictures corresponding to the initial picture. For example, for each first initial picture, a ratio of the number of pictures belonging to the first initial picture set in the corresponding first recalled picture in the first initial picture to the number of the first recalled pictures may be determined, where an embodiment refers to the ratio as a first ratio, and it is determined whether the first initial picture is noise according to the first ratio. The second initial picture is similar.
S380, if the first proportion is smaller than a fifth set threshold value, deleting the first initial picture from the first initial picture set; and if the second proportion is smaller than a sixth set threshold, deleting the second initial picture from the second initial picture set.
Specifically, if the first ratio is smaller than the fifth set threshold, the first initial picture is considered as noise, and the first initial picture may be deleted from the first initial picture set, and similarly, if the second ratio is smaller than the sixth set threshold, the second initial picture is considered as noise, and the second initial picture may be deleted from the second initial picture set, so that the first initial picture set and the second initial picture set may be optimized. The magnitudes of the fifth set threshold and the sixth set threshold may be set according to actual conditions.
The updating process can be repeatedly executed to continuously optimize the initial label system and meet the requirements of users.
Exemplarily, referring to fig. 4, fig. 4 is a schematic diagram of an implementation process of a tag hierarchy updating method based on a picture according to a third embodiment of the present disclosure. Specifically, the initial image set and the tags corresponding to the initial image set can be obtained by performing conventional clustering analysis on the acquired original images to form an initial tag system, and then the initial image set and the corresponding tags can be merged by the scheme, and the data of the initial image set can be cleaned to update the initial tag system.
On the basis of the embodiment, a recall picture corresponding to an initial picture can be determined from a picture library according to the initial picture in the initial picture set, similarity scores of two initial picture sets are determined according to the recall picture, and merging of the initial picture sets and merging of labels are further realized according to the similarity scores; on the basis, when the two initial picture sets cannot be combined, whether the corresponding initial picture is noise or not can be further determined according to the ratio of the number of the pictures belonging to the label to the number of the recalled pictures, and when the initial picture is noise, the initial picture is deleted from the initial picture set, so that the corresponding initial picture set is optimized, manual operation is not needed, the labor cost and the time cost are reduced, and the updating efficiency is improved.
Example four
Fig. 5 is a structural diagram of a device for updating a picture-based label system according to a fourth embodiment of the present disclosure, where the device may execute the method for updating a picture-based label system according to the foregoing embodiment, and referring to fig. 5, the device may include:
an obtaining module 31, configured to obtain an initial tag system to be updated, where the initial tag system includes at least two tags and an initial picture set corresponding to each tag, and the initial picture set includes at least one initial picture;
a recalled picture determining module 32, configured to determine, for a first initial picture set and a second initial picture set, a first recalled picture of a first initial picture in the first initial picture set and a second recalled picture of a second initial picture in the second initial picture set; the first initial picture set and the second initial picture set are any two picture sets in the initial picture sets, and the first recalled picture and the second recalled picture are pictures which are similar to the first initial picture and the second initial picture respectively in a picture library;
a similarity score determining module 33, configured to determine similarity scores of the first initial picture set and the second initial picture set according to the first recalled picture and the second recalled picture;
an updating module 34, configured to merge the first initial image set and the second initial image set and merge a first tag corresponding to the first initial image set and a second tag corresponding to the second initial image set if the similarity score meets a first preset condition, to obtain a target tag system, where the target tag system is used as an updating result of the initial tag system.
The fourth embodiment of the present disclosure provides a device for updating a tag system based on pictures, where an initial tag system to be updated is obtained, where the initial tag system includes at least two tags and an initial picture set corresponding to each tag, and the initial picture set includes at least one initial picture; determining a first recall picture of a first initial picture in the first initial picture set and a second recall picture of a second initial picture in the second initial picture set aiming at the first initial picture set and the second initial picture set; determining similarity scores of the first initial picture set and the second initial picture set according to the first recall picture and the second recall picture; and if the similarity score meets a first preset condition, merging the first initial image set and the second initial image set and merging a first label corresponding to the first initial image set and a second label corresponding to the second initial image set to obtain a target label system, wherein the target label system is used as an updating result of the initial label system. According to the scheme, the recall picture corresponding to the initial picture can be automatically determined according to the initial picture, the similarity scores of the two initial picture sets are determined according to the recall picture, when the similarity scores meet the first preset condition, the two initial picture sets and the corresponding labels are combined, the automatic updating of the initial label system is achieved, manual operation is not needed, the labor cost and the time cost are saved, and the updating efficiency is improved.
On the basis of the foregoing embodiment, the recall picture determining module 32 is specifically configured to:
respectively extracting a first picture feature of the first initial picture and a second picture feature of the second initial picture;
determining a first similarity score of the first picture feature and a reference picture feature and a second similarity score of the second picture feature and the reference picture feature, wherein the reference picture feature is a picture feature in the picture library, and the picture library is used for storing a corresponding relation between the reference picture feature and a picture;
if the first similarity score is larger than or equal to a first set threshold value, taking a picture corresponding to the reference picture characteristic as a first recalled picture; and if the second similarity score is larger than or equal to a second set threshold value, taking the picture corresponding to the reference picture characteristic as a second recalled picture.
On the basis of the above embodiment, the similarity score determining module 33 includes:
a number determining unit, configured to determine a first number of the first recalled pictures that belong to the second initial picture set and a second number of the second recalled pictures that belong to the first initial picture set;
a similarity score determining unit, configured to determine similarity scores of the first initial picture set and the second initial picture set according to the first number and the second number, and a first initial number of first initial pictures included in the first initial picture set and a second initial number of second initial pictures included in the second initial picture set.
On the basis of the foregoing embodiment, the similarity score determining unit is specifically configured to:
accumulating the first quantity and the second quantity to obtain a third quantity, and accumulating the first initial quantity and the second initial quantity to obtain a third initial quantity;
and taking the ratio of the third number to the third initial number as the similarity score of the first initial picture set and the second initial picture set.
On the basis of the above embodiment, the apparatus may further include:
a picture quantity determining module, configured to determine, if the similarity score does not satisfy a first preset condition, a first picture quantity that belongs to the first initial picture set in a first recalled picture corresponding to the first initial picture and a second picture quantity that belongs to the second initial picture set in a second recalled picture corresponding to the second initial picture;
a first deleting module, configured to delete the first initial picture from the first initial picture set if the number of the first pictures is smaller than a third set threshold; and if the number of the second pictures is smaller than a fourth set threshold value, deleting the second initial pictures from the second initial picture set.
On the basis of the above embodiment, the apparatus may further include:
a proportion determining module, configured to determine, if the similarity score does not satisfy a first preset condition, a first proportion of the number of pictures belonging to the first initial picture set in a first recalled picture corresponding to the first initial picture and the number of pictures of the first recalled picture, and a second proportion of the number of pictures belonging to the second initial picture set in a second recalled picture corresponding to the second initial picture and the number of pictures of the second recalled picture;
a second deleting module, configured to delete the first initial picture from the first initial picture set if the first percentage is smaller than a fifth set threshold; and if the second proportion is smaller than a sixth set threshold, deleting the second initial picture from the second initial picture set.
On the basis of the foregoing embodiment, the obtaining module 31 is specifically configured to:
acquiring an original picture;
extracting the characteristics of the original picture;
clustering the original pictures according to the characteristics to obtain an initial picture set and labels corresponding to the initial picture set;
and forming an initial label system according to the initial image set and the labels corresponding to the initial image set.
The image-based tag system updating apparatus provided by the embodiment of the present disclosure and the image-based tag system updating method provided by the above embodiment belong to the same inventive concept, and the technical details that are not described in detail in the embodiment can be referred to the above embodiment, and the embodiment has the same beneficial effects as the image-based tag system updating method.
EXAMPLE five
Referring now to FIG. 6, a block diagram of an electronic device 600 suitable for use in implementing embodiments of the present disclosure is shown. The electronic devices in the embodiments of the present disclosure may include, but are not limited to, mobile terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., car navigation terminals), and the like, and fixed terminals such as digital TVs, desktop computers, and servers, and the like. The electronic device shown in fig. 6 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. 6, electronic device 600 may include a processing means (e.g., central processing unit, graphics processor, etc.) 601 that may perform various appropriate actions and processes in accordance with a program stored in a Read Only Memory (ROM)602 or a program loaded from a storage means 608 into a Random Access Memory (RAM) 603. In the RAM 603, various programs and data necessary for the operation of the electronic apparatus 600 are also stored. The processing device 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input/output (I/O) interface 605 is also connected to bus 604.
Generally, the following devices may be connected to the I/O interface 605: input devices 606 including, for example, a touch screen, touch pad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; output devices 607 including, for example, a Liquid Crystal Display (LCD), a speaker, a vibrator, and the like; storage 608 including, for example, tape, hard disk, etc.; and a communication device 609. The communication means 609 may allow the electronic device 600 to communicate with other devices wirelessly or by wire to exchange data. While fig. 6 illustrates an electronic device 600 having various means, it is to be understood that not all illustrated means are required to be implemented or provided. More or fewer devices may alternatively be implemented or provided.
In particular, according to an embodiment of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program carried on a non-transitory computer readable medium, the computer program containing program code for performing the method illustrated by the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network via the communication means 609, or may be installed from the storage means 608, or may be installed from the ROM 602. The computer program, when executed by the processing device 601, performs the above-described functions defined in the methods of the embodiments of the present disclosure.
EXAMPLE six
The computer readable medium described above in this disclosure 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 comprise a propagated data signal with computer readable program code embodied therein, either 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: electrical wires, optical cables, RF (radio frequency), etc., or any suitable combination of the foregoing.
In some embodiments, the clients, servers may communicate using any currently known or future developed network Protocol, such as HTTP (HyperText Transfer Protocol), and may interconnect with any form or medium of digital data communication (e.g., a communications network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), the Internet (e.g., the Internet), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future developed network.
The computer readable medium may be embodied in the electronic device; or may exist separately without being assembled into the electronic device.
The computer readable medium carries one or more programs which, when executed by the electronic device, cause the electronic device to: acquiring an initial label system to be updated, wherein the initial label system comprises at least two labels and an initial picture set corresponding to each label, and the initial picture set comprises at least one initial picture; determining a first recall picture of a first initial picture in the first initial picture set and a second recall picture of a second initial picture in the second initial picture set aiming at the first initial picture set and the second initial picture set; the first initial picture set and the second initial picture set are any two picture sets in the initial picture sets, and the first recalled picture and the second recalled picture are pictures which are similar to the first initial picture and the second initial picture respectively in a picture library; determining similarity scores of the first initial picture set and the second initial picture set according to the first recall picture and the second recall picture; and if the similarity score meets a first preset condition, merging the first initial image set and the second initial image set and merging a first label corresponding to the first initial image set and a second label corresponding to the second initial image set to obtain a target label system, wherein the target label system is used as an updating result of the initial label system.
Computer program code for carrying out operations for the present disclosure may be written in any combination of one or more programming languages, including but not limited to an object oriented programming language such as Java, Smalltalk, C + +, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).
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 and/or flowchart illustration, and combinations of blocks in the block diagrams and/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.
The modules described in the embodiments of the present disclosure may be implemented by software or hardware. The name of a module does not in some cases form a limitation on the module itself, and for example, the obtaining module may also be described as a module for obtaining an initial label system to be updated.
The functions described herein above may be performed, at least in part, by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), systems on a chip (SOCs), Complex Programmable Logic Devices (CPLDs), and the like.
In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on 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.
According to one or more embodiments of the present disclosure, the present disclosure provides a method for updating a label hierarchy based on a picture, including:
acquiring an initial label system to be updated, wherein the initial label system comprises at least two labels and an initial picture set corresponding to each label, and the initial picture set comprises at least one initial picture;
determining a first recall picture of a first initial picture in the first initial picture set and a second recall picture of a second initial picture in the second initial picture set aiming at the first initial picture set and the second initial picture set; the first initial picture set and the second initial picture set are any two picture sets in the initial picture sets, and the first recalled picture and the second recalled picture are pictures which are similar to the first initial picture and the second initial picture respectively in a picture library;
determining similarity scores of the first initial picture set and the second initial picture set according to the first recall picture and the second recall picture;
and if the similarity score meets a first preset condition, merging the first initial image set and the second initial image set and merging a first label corresponding to the first initial image set and a second label corresponding to the second initial image set to obtain a target label system, wherein the target label system is used as an updating result of the initial label system.
According to one or more embodiments of the present disclosure, in the method for updating a picture-based tag system provided by the present disclosure, the determining a first recalled picture of a first initial picture in the first initial picture set and a second recalled picture of a second initial picture in the second initial picture set includes:
respectively extracting a first picture feature of the first initial picture and a second picture feature of the second initial picture;
determining a first similarity score of the first picture feature and a reference picture feature and a second similarity score of the second picture feature and the reference picture feature, wherein the reference picture feature is a picture feature in the picture library, and the picture library is used for storing a corresponding relation between the reference picture feature and a picture;
if the first similarity score is larger than or equal to a first set threshold value, taking a picture corresponding to the reference picture characteristic as a first recalled picture; and if the second similarity score is larger than or equal to a second set threshold value, taking the picture corresponding to the reference picture characteristic as a second recalled picture.
According to one or more embodiments of the present disclosure, in the method for updating a picture-based tag system provided by the present disclosure, determining similarity scores of the first initial picture set and the second initial picture set according to the first recalled picture and the second recalled picture includes:
determining a first number of the first recalled pictures that belong to the second initial picture set and a second number of the second recalled pictures that belong to the first initial picture set;
determining similarity scores of the first initial picture set and the second initial picture set according to the first number and the second number, and a first initial number of first initial pictures contained in the first initial picture set and a second initial number of second initial pictures contained in the second initial picture set.
According to one or more embodiments of the present disclosure, in the method for updating a picture-based label system provided by the present disclosure, determining similarity scores of a first initial picture set and a second initial picture set according to the first number and the second number and a first initial number of first initial pictures included in the first initial picture set and a second initial number of second initial pictures included in the second initial picture set includes:
accumulating the first quantity and the second quantity to obtain a third quantity, and accumulating the first initial quantity and the second initial quantity to obtain a third initial quantity;
and taking the ratio of the third number to the third initial number as the similarity score of the first initial picture set and the second initial picture set.
According to one or more embodiments of the present disclosure, the method for updating a label hierarchy based on pictures provided by the present disclosure further includes:
if the similarity score does not meet a first preset condition, determining a first picture quantity belonging to the first initial picture set in a first recalled picture corresponding to the first initial picture and a second picture quantity belonging to the second initial picture set in a second recalled picture corresponding to the second initial picture;
if the number of the first pictures is smaller than a third set threshold value, deleting the first initial pictures from the first initial picture set; and if the number of the second pictures is smaller than a fourth set threshold value, deleting the second initial pictures from the second initial picture set.
According to one or more embodiments of the present disclosure, the method for updating a label hierarchy based on pictures provided by the present disclosure further includes:
if the similarity score does not meet a first preset condition, determining a first ratio of the number of pictures belonging to the first initial picture set in a first recalled picture corresponding to the first initial picture to the number of pictures belonging to the first recalled picture, and a second ratio of the number of pictures belonging to the second initial picture set in a second recalled picture corresponding to the second initial picture to the number of pictures belonging to the second recalled picture;
deleting the first initial picture from the first initial picture set if the first proportion is smaller than a fifth set threshold; and if the second proportion is smaller than a sixth set threshold, deleting the second initial picture from the second initial picture set.
According to one or more embodiments of the present disclosure, in the method for updating a tag hierarchy based on a picture provided by the present disclosure, the obtaining of an initial tag hierarchy to be updated includes:
acquiring an original picture;
extracting the characteristics of the original picture;
clustering the original pictures according to the characteristics to obtain an initial picture set and labels corresponding to the initial picture set;
and forming an initial label system according to the initial image set and the labels corresponding to the initial image set.
According to one or more embodiments of the present disclosure, the present disclosure provides a tag hierarchy updating apparatus based on a picture, including:
the system comprises an acquisition module, a updating module and a display module, wherein the acquisition module is used for acquiring an initial label system to be updated, the initial label system comprises at least two labels and an initial picture set corresponding to each label, and the initial picture set comprises at least one initial picture;
a recalled picture determining module, configured to determine, for a first initial picture set and a second initial picture set, a first recalled picture of a first initial picture in the first initial picture set and a second recalled picture of a second initial picture in the second initial picture set; the first initial picture set and the second initial picture set are any two picture sets in the initial picture sets, and the first recalled picture and the second recalled picture are pictures which are similar to the first initial picture and the second initial picture respectively in a picture library;
a similarity score determining module, configured to determine similarity scores of the first initial picture set and the second initial picture set according to the first recalled picture and the second recalled picture;
and the updating module is used for merging the first initial image set and the second initial image set and merging a first label corresponding to the first initial image set and a second label corresponding to the second initial image set to obtain a target label system as an updating result of the initial label system if the similarity score meets a first preset condition.
In accordance with one or more embodiments of the present disclosure, there is provided an electronic device including:
one or more processors;
a memory for storing one or more programs;
the one or more programs, when executed by the one or more processors, implement any of the picture-based label hierarchy update methods provided in this disclosure.
According to one or more embodiments of the present disclosure, there is provided a computer-readable storage medium having stored thereon a computer program, which when executed by a processor, implements any one of the image-based label system updating methods provided by the present disclosure.
The foregoing description is only exemplary of the preferred embodiments of the disclosure and is illustrative of the principles of the technology employed. It will be appreciated by those skilled in the art that the scope of the disclosure herein is not limited to the particular combination of features described above, but also encompasses other embodiments in which any combination of the features described above or their equivalents does not depart from the spirit of the disclosure. For example, the above features and (but not limited to) the features disclosed in this disclosure having similar functions are replaced with each other to form the technical solution.
Further, while operations are depicted in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Likewise, while several specific implementation details are included in the above discussion, these should not be construed as limitations on the scope of the disclosure. Certain features that are described in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination.
Although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims.
Claims (10)
1. A label system updating method based on pictures is characterized by comprising the following steps:
acquiring an initial label system to be updated, wherein the initial label system comprises at least two labels and an initial picture set corresponding to each label, and the initial picture set comprises at least one initial picture;
determining a first recall picture of a first initial picture in the first initial picture set and a second recall picture of a second initial picture in the second initial picture set aiming at the first initial picture set and the second initial picture set; the first initial picture set and the second initial picture set are any two picture sets in the initial picture sets, and the first recalled picture and the second recalled picture are pictures which are similar to the first initial picture and the second initial picture respectively in a picture library;
determining similarity scores of the first initial picture set and the second initial picture set according to the first recall picture and the second recall picture;
and if the similarity score meets a first preset condition, merging the first initial image set and the second initial image set and merging a first label corresponding to the first initial image set and a second label corresponding to the second initial image set to obtain a target label system, wherein the target label system is used as an updating result of the initial label system.
2. The method of claim 1, wherein the determining a first recalled picture of a first initial picture of the first initial set of pictures and a second recalled picture of a second initial picture of the second initial set of pictures comprises:
respectively extracting a first picture feature of the first initial picture and a second picture feature of the second initial picture;
determining a first similarity score of the first picture feature and a reference picture feature and a second similarity score of the second picture feature and the reference picture feature, wherein the reference picture feature is a picture feature in the picture library, and the picture library is used for storing a corresponding relation between the reference picture feature and a picture;
if the first similarity score is larger than or equal to a first set threshold value, taking a picture corresponding to the reference picture characteristic as a first recalled picture; and if the second similarity score is larger than or equal to a second set threshold value, taking the picture corresponding to the reference picture characteristic as a second recalled picture.
3. The method of claim 1, wherein the determining similarity scores for the first initial set of pictures and the second initial set of pictures from the first recalled picture and the second recalled picture comprises:
determining a first number of the first recalled pictures that belong to the second initial picture set and a second number of the second recalled pictures that belong to the first initial picture set;
determining similarity scores of the first initial picture set and the second initial picture set according to the first number and the second number, and a first initial number of first initial pictures contained in the first initial picture set and a second initial number of second initial pictures contained in the second initial picture set.
4. The method of claim 3, wherein determining the similarity scores of the first initial picture set and the second initial picture set according to the first number and the second number and a first initial number of first initial pictures included in the first initial picture set and a second initial number of second initial pictures included in the second initial picture set comprises:
accumulating the first quantity and the second quantity to obtain a third quantity, and accumulating the first initial quantity and the second initial quantity to obtain a third initial quantity;
and taking the ratio of the third number to the third initial number as the similarity score of the first initial picture set and the second initial picture set.
5. The method of claim 1, further comprising:
if the similarity score does not meet a first preset condition, determining a first picture quantity belonging to the first initial picture set in a first recalled picture corresponding to the first initial picture and a second picture quantity belonging to the second initial picture set in a second recalled picture corresponding to the second initial picture;
if the number of the first pictures is smaller than a third set threshold value, deleting the first initial pictures from the first initial picture set; and if the number of the second pictures is smaller than a fourth set threshold value, deleting the second initial pictures from the second initial picture set.
6. The method of claim 1, further comprising:
if the similarity score does not meet a first preset condition, determining a first ratio of the number of pictures belonging to the first initial picture set in a first recalled picture corresponding to the first initial picture to the number of pictures belonging to the first recalled picture, and a second ratio of the number of pictures belonging to the second initial picture set in a second recalled picture corresponding to the second initial picture to the number of pictures belonging to the second recalled picture;
deleting the first initial picture from the first initial picture set if the first proportion is smaller than a fifth set threshold; and if the second proportion is smaller than a sixth set threshold, deleting the second initial picture from the second initial picture set.
7. The method according to any one of claims 1-6, wherein the obtaining the initial label hierarchy to be updated comprises:
acquiring an original picture;
extracting the characteristics of the original picture;
clustering the original pictures according to the characteristics to obtain an initial picture set and labels corresponding to the initial picture set;
and forming an initial label system according to the initial image set and the labels corresponding to the initial image set.
8. A label system updating device based on pictures is characterized by comprising:
the system comprises an acquisition module, a updating module and a display module, wherein the acquisition module is used for acquiring an initial label system to be updated, the initial label system comprises at least two labels and an initial picture set corresponding to each label, and the initial picture set comprises at least one initial picture;
a recalled picture determining module, configured to determine, for a first initial picture set and a second initial picture set, a first recalled picture of a first initial picture in the first initial picture set and a second recalled picture of a second initial picture in the second initial picture set; the first initial picture set and the second initial picture set are any two picture sets in the initial picture sets, and the first recalled picture and the second recalled picture are pictures which are similar to the first initial picture and the second initial picture respectively in a picture library;
a similarity score determining module, configured to determine similarity scores of the first initial picture set and the second initial picture set according to the first recalled picture and the second recalled picture;
and the updating module is used for merging the first initial image set and the second initial image set and merging a first label corresponding to the first initial image set and a second label corresponding to the second initial image set to obtain a target label system as an updating result of the initial label system if the similarity score meets a first preset condition.
9. An electronic device, comprising:
one or more processors;
a memory for storing one or more programs;
the one or more programs when executed by the one or more processors implement the picture-based labeling architecture updating method of any of claims 1-7.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the picture-based label system updating method according to any one of claims 1 to 7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011360402.4A CN112395447A (en) | 2020-11-27 | 2020-11-27 | Label system updating method, device and equipment based on picture and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011360402.4A CN112395447A (en) | 2020-11-27 | 2020-11-27 | Label system updating method, device and equipment based on picture and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112395447A true CN112395447A (en) | 2021-02-23 |
Family
ID=74604652
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011360402.4A Pending CN112395447A (en) | 2020-11-27 | 2020-11-27 | Label system updating method, device and equipment based on picture and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112395447A (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108062377A (en) * | 2017-12-12 | 2018-05-22 | 百度在线网络技术(北京)有限公司 | The foundation of label picture collection, definite method, apparatus, equipment and the medium of label |
CN109857706A (en) * | 2019-01-16 | 2019-06-07 | 深圳壹账通智能科技有限公司 | Picture method for cleaning, device and the terminal device updated based on data |
CN110597719A (en) * | 2019-09-05 | 2019-12-20 | 腾讯科技(深圳)有限公司 | Image clustering method, device and medium for adaptation test |
CN111414949A (en) * | 2020-03-13 | 2020-07-14 | 杭州海康威视***技术有限公司 | Picture clustering method and device, electronic equipment and storage medium |
-
2020
- 2020-11-27 CN CN202011360402.4A patent/CN112395447A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108062377A (en) * | 2017-12-12 | 2018-05-22 | 百度在线网络技术(北京)有限公司 | The foundation of label picture collection, definite method, apparatus, equipment and the medium of label |
CN109857706A (en) * | 2019-01-16 | 2019-06-07 | 深圳壹账通智能科技有限公司 | Picture method for cleaning, device and the terminal device updated based on data |
CN110597719A (en) * | 2019-09-05 | 2019-12-20 | 腾讯科技(深圳)有限公司 | Image clustering method, device and medium for adaptation test |
CN111414949A (en) * | 2020-03-13 | 2020-07-14 | 杭州海康威视***技术有限公司 | Picture clustering method and device, electronic equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110634047B (en) | Method and device for recommending house resources, electronic equipment and storage medium | |
CN110413812B (en) | Neural network model training method and device, electronic equipment and storage medium | |
CN112650790B (en) | Target point cloud plane determining method and device, electronic equipment and storage medium | |
CN111784712B (en) | Image processing method, device, equipment and computer readable medium | |
CN110991373A (en) | Image processing method, image processing apparatus, electronic device, and medium | |
CN112364829B (en) | Face recognition method, device, equipment and storage medium | |
CN114494709A (en) | Feature extraction model generation method, image feature extraction method and device | |
CN111241137B (en) | Data processing method, device, electronic equipment and storage medium | |
CN111209432A (en) | Information acquisition method and device, electronic equipment and computer readable medium | |
CN110633383A (en) | Method and device for identifying repeated house sources, electronic equipment and readable medium | |
CN113033682B (en) | Video classification method, device, readable medium and electronic equipment | |
CN111414921B (en) | Sample image processing method, device, electronic equipment and computer storage medium | |
CN113220281A (en) | Information generation method and device, terminal equipment and storage medium | |
CN112380883A (en) | Model training method, machine translation method, device, equipment and storage medium | |
CN115269978A (en) | Video tag generation method, device, equipment and medium | |
CN112395447A (en) | Label system updating method, device and equipment based on picture and storage medium | |
CN111737575B (en) | Content distribution method, content distribution device, readable medium and electronic equipment | |
CN110413603B (en) | Method and device for determining repeated data, electronic equipment and computer storage medium | |
CN114187557A (en) | Method, device, readable medium and electronic equipment for determining key frame | |
CN112417276A (en) | Paging data acquisition method and device, electronic equipment and computer readable storage medium | |
CN112488947A (en) | Model training and image processing method, device, equipment and computer readable medium | |
CN111680754A (en) | Image classification method and device, electronic equipment and computer-readable storage medium | |
CN111368204A (en) | Content pushing method and device, electronic equipment and computer readable medium | |
CN110716885B (en) | Data management method and device, electronic equipment and storage medium | |
CN114040014B (en) | Content pushing method, device, electronic equipment and computer readable storage medium |
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 |