CN114332477A - Feature recognition model training method, item feature recognition method and device - Google Patents

Abstract

The present disclosure provides a feature recognition model training method, an item feature recognition method and device. The training method includes: acquiring an item sample data set, the item sample data set includes a plurality of item samples, and inputting the item image data, text data used to describe the item and item style data in each item sample into the image features of the initial model respectively The extraction layer, text feature extraction layer and style feature extraction layer output the item image features, text features and style features respectively; input the item image features and text features into the feature combination layer of the initial model, and output the item combination features; The style feature is input into the matching layer of the initial model, and the matching result is output; the model parameters of the initial model are adjusted according to the matching result and the style label, and the trained feature recognition model is obtained.

Description

Feature recognition model training method, item feature recognition method and device

technical field

The present disclosure relates to the field of computer technologies, and more particularly, to a method for training a feature recognition model, a method and apparatus for recognizing item features, an electronic device, a computer-readable storage medium, and a computer program product.

Background technique

The traditional method of item style labeling mainly uses manual labeling. This method requires a large number of manual editing rules to be set in advance, and there are problems such as labor consumption, material resources, low labeling efficiency, and poor versatility. Moreover, in the process of setting manual editing rules , due to differences in human subjective cognition, it will lead to differences in the setting of editing rules, which in turn lead to differences in the results of manual annotation styles.

In the related art, a machine learning algorithm is used to realize the labeling of item styles, but this machine learning algorithm has poor generalization to unfamiliar items and unfamiliar styles.

SUMMARY OF THE INVENTION

In view of this, the present disclosure provides a feature recognition model training method, an item feature recognition method and apparatus, an electronic device, a computer-readable storage medium, and a computer program product.

A first aspect of the present disclosure provides a feature recognition model training method, including:

obtaining an item sample data set, wherein the item sample data set includes a plurality of item samples, and each item sample includes item image data, text data for describing the item, and item style data, wherein the item sample has a style label;

For each item sample, input the item image data into the image feature extraction layer of the initial model, and output the item image features; input the text data used to describe the item into the text feature extraction layer of the initial model, and output the text features; input the item style data into The style feature extraction layer of the initial model outputs style features;

Input the item image features and text features into the feature combination layer of the initial model, and output the item combination features;

Input the item combination feature and style feature into the matching layer of the initial model, and output the matching result used to characterize the item combination feature and style feature;

The model parameters of the initial model are adjusted according to the matching results and style labels, and the trained feature recognition model is obtained.

According to an embodiment of the present disclosure, the item image features and text features are input into the image combination layer of the initial model, and the output item combination features include:

The image combination layer is used to stitch item image features and text features into item combination features.

According to an embodiment of the present disclosure, after acquiring the item sample data set, the method further includes:

Generate an augmented item sample dataset from the item sample dataset.

According to an embodiment of the present disclosure, generating an augmented item sample data set according to the item sample data set includes:

According to the item image data, from the item database, determine a first item data list that is similar to the item image data, wherein the first item data list includes item image data of different items, text data for describing the item, and item style data;

According to the text data for describing the item, from the item database, a second item data list similar to the text data for describing the item is determined, wherein the second item data list includes item image data for different items, and a second item data list for describing the item. text data and item style data;

An augmented item sample data set is generated according to the first item data list and the second item data list.

According to an embodiment of the present disclosure, the model parameters of the initial model include model parameters of the image feature extraction layer of the initial model, model parameters of the text feature extraction layer of the initial model, and model parameters of the style feature extraction layer of the initial model. According to the matching result and The style label adjusts the model parameters of the initial model to obtain the trained feature recognition model, including:

According to the matching results and style labels, the model parameters of the image feature extraction layer of the initial model, the model parameters of the text feature extraction layer of the initial model, and the model parameters of the style feature extraction layer of the initial model are adjusted to obtain the trained feature recognition model.

A second aspect of the present disclosure provides an item feature identification method, comprising:

Obtain item information of the item to be processed, wherein the item information includes item image information and text information used to describe the item;

Input the image information of the item into the image feature extraction layer of the feature recognition model, and output the item image feature of the item to be processed, wherein the feature recognition model is trained by the training method of the embodiment of the present disclosure;

Input the text information used to describe the item into the text feature extraction layer of the feature recognition model, and output the text feature of the item to be processed;

Input the image feature and text feature of the item into the feature combination layer of the feature recognition model, and output the item combination feature of the item to be processed;

The item combination feature and the description feature of the candidate style are input into the matching layer of the feature recognition model, and the matching result used to characterize the item combination feature and the style feature is output, wherein the description feature of the candidate style is the item style obtained from the candidate style database. The information is obtained after inputting the style feature extraction layer of the feature recognition model;

According to the matching result, the style feature information matching the item combination feature of the item to be processed is determined.

According to an embodiment of the present disclosure, the above feature identification method further includes:

Get multiple item review text data and multiple item title text data;

Preprocessing multiple item review text data and multiple item title text data to obtain a text data set for characterizing item style;

Generate a database of candidate styles from a text dataset used to characterize item styles.

According to an embodiment of the present disclosure, the above feature identification method further includes:

Generate a vector dataset for characterizing item style from a text dataset for characterizing item style;

Generate a candidate style database from the vector dataset used to characterize item styles.

A third aspect of the present disclosure provides a feature recognition model training device, including: a first acquisition module, a feature extraction module, a feature combination module, a matching module, and an adjustment module. The first acquisition module is used to acquire an item sample data set, wherein the item sample data set includes a plurality of item samples, and each item sample includes item image data, text data for describing the item, and item style data, and the item sample Has style labels. The feature extraction module is used to input the item image data into the image feature extraction layer of the initial model for each item sample, and output the item image features; input the text data used to describe the item into the text feature extraction layer of the initial model, and output the text features ; Input the item style data into the style feature extraction layer of the initial model, and output the style feature. The feature combination module is used to input the item image features and text features into the feature combination layer of the initial model, and output the item combination features. The matching module is used to input the item combination feature and style feature into the matching layer of the initial model, and output the matching result used to characterize the item combination feature and style feature. The adjustment module is used to adjust the model parameters of the initial model according to the matching results and style labels, and obtain the trained feature recognition model.

According to an embodiment of the present disclosure, the feature combination module includes a stitching unit for stitching the item image feature and the text feature into the item combination feature using the image combination layer.

According to an embodiment of the present disclosure, the above-mentioned apparatus further includes: a generating module configured to generate an augmented item sample data set according to the item sample data set.

According to an embodiment of the present disclosure, the generating module includes a first determining unit, a second determining unit, and a generating unit. The first determining unit determines a first item data list similar to the item image data from the item database according to the item image data, wherein the first item data list includes item image data of different items, a description of the item for describing the item Text data and item style data. The second determining unit is configured to determine a second item data list similar to the text data for describing the item from the item database according to the text data for describing the item, wherein the second item data list includes items of different items Image data, text data to describe the item, and item style data. The generating unit is configured to generate an augmented item sample data set according to the first item data list and the second item data list.

According to an embodiment of the present disclosure, the generating unit includes a generating subunit for generating an augmented item sample data set according to the data intersection of the first item data list and the second item data list.

According to an embodiment of the present disclosure, the adjustment module includes an adjustment unit for adjusting the model parameters of the image feature extraction layer of the initial model, the model parameters of the text feature extraction layer of the initial model, and the style features of the initial model according to the matching result and the style label Extract the model parameters of the layer to obtain the trained feature recognition model.

A fourth aspect of the present disclosure provides an article feature identification device, comprising: a second acquisition module, a first identification module, a second identification module, a third identification module, a fourth identification module, and a determination module. Wherein, the second acquisition module is used for acquiring item information of the item to be processed, wherein the item information includes item image information and text information for describing the item. The first recognition module is used for inputting the image information of the item into the image feature extraction layer of the feature recognition model, and outputting the item image feature of the item to be processed, wherein the feature recognition model is obtained by training the training method provided by the embodiment of the present disclosure. The second recognition module is configured to input the text information for describing the item into the text feature extraction layer of the feature recognition model, and output the text feature of the item to be processed. The third recognition module is used for inputting the image feature of the item and the text feature into the feature combination layer of the feature recognition model, and outputting the item combination feature of the item to be processed. a fourth recognition module, configured to input the item combination feature and the description feature of the candidate style into the matching layer of the feature recognition model, and output a matching result used to characterize the item combination feature and the description feature of the candidate style, Wherein, the description feature of the candidate style is obtained by inputting the item style information obtained from the candidate style database into the style feature extraction layer of the feature recognition model. A determination module, configured to determine, according to the matching result, style feature information matching the item combination feature of the item to be processed.

A fifth aspect of the present disclosure provides an electronic device, comprising: one or more processors; a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more programs When executed by the processor, one or more processors are caused to execute the above feature recognition model training method.

A sixth aspect of the present disclosure also provides a computer-readable storage medium on which executable instructions are stored, and when the instructions are executed by a processor, cause the processor to execute the above feature recognition model training method.

A seventh aspect of the present disclosure also provides a computer program product, including a computer program, which implements the above feature recognition model training method when the computer program is executed by a processor.

According to the embodiment of the present disclosure, since the item sample data set including item image data, text data for describing the item, item style data, and style tags is used, the item image data and the text data for describing the item are respectively input by inputting the item image data and the text data for describing the item. The image feature extraction layer and text feature extraction layer of the initial model output image features and text features, and then input the feature combination layer to output the item combination feature, and input the item combination feature and the style feature extracted by the style feature extraction layer of the initial model into the matching layer. The feature recognition model is obtained by training the technical means of adjusting the model parameters of the initial model through the output matching results and style labels, so at least partially overcomes the technical problem of poor generalization of unfamiliar objects and styles using machine learning algorithms in the related art, and improves the Generalizability to Unfamiliar Object and Unfamiliar Style Recognition.

Description of drawings

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

FIG. 1 schematically shows an exemplary system architecture to which the feature recognition model training method according to the embodiment of the present disclosure can be applied;

FIG. 2 schematically shows a flowchart of a feature recognition model training method according to an embodiment of the present disclosure;

FIG. 3 schematically shows a flowchart of a method for amplifying a sample data set according to an embodiment of the present disclosure;

FIG. 4 schematically shows a schematic diagram of the architecture of a feature recognition model according to an embodiment of the present disclosure;

FIG. 5 schematically shows a flowchart of a method for identifying an item feature according to an embodiment of the present disclosure;

6 schematically shows a flowchart of a method for generating a candidate style database according to an embodiment of the present disclosure;

FIG. 7 schematically shows a system architecture diagram of item feature identification according to an embodiment of the present disclosure;

8 schematically shows a block diagram of a feature recognition model training apparatus according to an embodiment of the present disclosure;

FIG. 9 schematically shows a block diagram of an apparatus for identifying characteristics of an item according to an embodiment of the present disclosure; and

FIG. 10 schematically shows a block diagram of an electronic device suitable for implementing a feature recognition model training method according to an embodiment of the present disclosure.

Detailed ways

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

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

All terms (including technical and scientific terms) used herein have the meaning as commonly understood by one of ordinary skill in the art, unless otherwise defined. It should be noted that the terms used herein should be construed to have meanings consistent with the context of the present specification and should not be construed in an idealized or overly rigid manner.

Where expressions like "at least one of A, B, and C, etc.," are used, they should generally be interpreted in accordance with the meaning of the expression as commonly understood by those skilled in the art (eg, "has A, B, and C") At least one of the "systems" shall include, but not be limited to, systems with A alone, B alone, C alone, A and B, A and C, B and C, and/or A, B, C, etc. ).

Embodiments of the present disclosure provide a method for training a feature recognition model, using an item sample data set including item image data, text data for describing items, item style data, and item style labels, by using item image data, used for The text data describing the item is input into the image feature extraction layer and text feature extraction layer of the initial model respectively to output image features and text features, and then input to the feature combination layer to output the item combination feature. The item combination feature and the style feature extraction layer using the initial model are extracted The style features are input to the matching layer, and the feature recognition model is obtained by training the technical means of adjusting the model parameters of the initial model through the output matching results and style labels.

FIG. 1 schematically illustrates an exemplary system architecture 100 to which a feature recognition model training method may be applied, according to an embodiment of the present disclosure. It should be noted that FIG. 1 is only an example of a system architecture to which the embodiments of the present disclosure can be applied, so as to help those skilled in the art to understand the technical content of the present disclosure, but it does not mean that the embodiments of the present disclosure cannot be used for other A device, system, environment or scene.

As shown in FIG. 1 , the system architecture 100 according to this embodiment may include terminal devices 101 , 102 , and 103 , a network 104 and a server 105 . The network 104 is a medium used to provide a communication link between the terminal devices 101 , 102 , 103 and the server 105 . The network 104 may include various connection types, such as wired and/or wireless communication links, and the like.

The user can use the terminal devices 101, 102, 103 to interact with the server 105 through the network 104 to receive or send messages and the like. Various communication client applications may be installed on the terminal devices 101, 102, and 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients and/or social platform software, etc. (just an example). ).

The terminal devices 101, 102, 103 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like.

The server 105 may be a server that provides various services, such as a background management server (just an example) that provides support for websites browsed by users using the terminal devices 101 , 102 , and 103 . The background management server can analyze and process the received user requests and other data, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal device.

It should be noted that, the feature recognition model training method or the feature recognition method provided by the embodiments of the present disclosure may generally be executed by the server 105 . Correspondingly, the feature recognition model training apparatus or the feature recognition apparatus provided by the embodiments of the present disclosure may generally be provided in the server 105 . The feature recognition model training method or feature recognition method provided by the embodiments of the present disclosure may also be executed by a server or server cluster different from the server 105 and capable of communicating with the terminal devices 101 , 102 , 103 and/or the server 105 . Correspondingly, the feature recognition model training apparatus or feature recognition apparatus provided by the embodiments of the present disclosure may also be provided in a server or server cluster different from the server 105 and capable of communicating with the terminal devices 101 , 102 , 103 and/or the server 105 . Alternatively, the feature recognition model training method or feature recognition method provided by the embodiments of the present disclosure may also be executed by the terminal device 101 , 102 , or 103 , or may also be executed by other terminal devices different from the terminal device 101 , 102 , or 103 . . Correspondingly, the feature recognition model training apparatus or feature recognition apparatus provided in the embodiments of the present disclosure may also be set in the terminal equipment 101 , 102 , or 103 , or set in other terminal equipment different from the terminal equipment 101 , 102 , or 103 . middle.

For example, the item sample data set can be acquired through any one of the terminal devices 101 , 102 , or 103 (for example, the terminal device 101 , but not limited to), and can also be stored in any one of the terminal devices 101 , 102 , or 103 or stored on an external storage device and can be imported into the terminal device 101 . Then, the terminal device 101 can locally execute the feature recognition model training method provided by the embodiment of the present disclosure, or send the item sample data set to other terminal devices, servers, or server clusters, and other devices that receive the item sample data set A terminal device, a server, or a server cluster is used to execute the feature recognition model training method provided by the embodiments of the present disclosure.

It should be understood that the numbers of terminal devices, networks and servers in FIG. 1 are merely illustrative. There can be any number of terminal devices, networks and servers according to implementation needs.

FIG. 2 schematically shows a flowchart of a method for training a feature recognition model according to an embodiment of the present disclosure.

As shown in FIG. 2, the feature recognition model training method of this embodiment includes operations S210-S270.

In operation S210, an item sample data set is obtained, wherein the item sample data set includes a plurality of item samples, and each item sample includes item image data, text data for describing the item, and item style data, wherein the item sample has a style tag .

According to an embodiment of the present disclosure, the image data of the item may include a picture of the item, and the image of the item may also include a model or a display stand, a display cabinet, and the like for displaying the item after the item itself. The text data used to describe the item may include text data describing the color of the item, text data describing the brand of the item, text data describing the style of the item, etc., for example: red pleated skirt, blue T-shirt of a certain sports brand, etc. The item style data may include sports style, casual style, business style, sweet style, and the like. The style tag can use a binary number to indicate whether the item matches the style. When the item matches the style, the style tag is 1; when the item does not match the style, the style tag is 0. For example, if the style of a blue T-shirt of a sports brand is sports style, the style label is 1. The style of the blue T-shirt of a sports brand is lady style, then the style tag is 0.

In operation S220, for each item sample, the item image data is input into the image feature extraction layer of the initial model, and the item image features are output.

According to an embodiment of the present disclosure, for example: the item image data is a picture of a blue T-shirt of a certain sports brand, the item image data is input into the image feature extraction layer of the initial model, and the output item image features may include blue, sports brand logos, T-shirts, etc.

In operation S230, the text data for describing the item is input into the text feature extraction layer of the initial model, and the text features are output.

According to an embodiment of the present disclosure, for example: the text data used to describe the item is a blue T-shirt of a certain sports brand, the item image data is input into the text data extraction layer of the initial model, and the output text data features may include blue, Sports brand names, t-shirts, etc.

In operation S240, the item style data is input into the style feature extraction layer of the initial model, and the style feature is output.

According to an embodiment of the present disclosure, for example, if the style of the blue T-shirt of a certain sports brand is sports style, the item style tag is 1, indicating that the item matches the style, and the sample data is positive sample data. Input the item style data into the style feature extraction layer of the initial model, and the output style feature can be motion.

In operation S250, the item image feature and the text feature are input into the feature combination layer of the initial model, and the item combination feature is output.

According to an embodiment of the present disclosure, the image feature "blue, the logo of a sports brand, T-shirt" and the text feature "blue, the name of a sports brand, T-shirt" of the above-mentioned items are input into the feature combination layer, and the output item combination feature It is "blue, the logo of the sports brand, the name of the sports brand, the T-shirt".

In operation S260, the item combination feature and the style feature are input into the matching layer of the initial model, and a matching result for characterizing the item combination feature and the style feature is output.

According to an embodiment of the present disclosure, the item combination feature "blue, the logo of the sports brand, the name of the sports brand, T-shirt" and the style feature "sports" are input into the matching layer of the initial model, and the output is used to characterize the item combination feature and style. The matching result of the feature, for example, the matching result is 0.8.

According to the embodiment of the present disclosure, the item combination feature and the style feature can be represented by a vector, and the matching layer of the initial model can determine the distance between the item combination feature vector and the style feature vector by calculating the distance between the item combination feature vector and the style feature vector, such as Euclidean distance, cosine similarity, etc. Matching results for characterizing item combination features and style features.

In operation S270, the model parameters of the initial model are adjusted according to the matching result and the style label, so as to obtain a trained feature recognition model.

According to the embodiment of the present disclosure, since the above input is positive sample data and the style label is 1, the model parameters of the initial model can be adjusted according to the matching result 0.8 and the style label 1. After the adjustment is completed, the output matching result can be changed to A value of 0.9, 0.95 or closer to 1, the closer the output matching result is to 1, the higher the recognition accuracy of the feature recognition model.

According to the embodiment of the present disclosure, since the item sample data set including item image data, text data for describing the item, item style data, and item style label is adopted, by dividing the item image data and the text data for describing the item, respectively Input the image feature extraction layer and text feature extraction layer of the initial model to output image features and text features, then input the feature combination layer to output the item combination feature, and input the item combination feature and the style feature extracted by the style feature extraction layer of the initial model into the matching layer. , through the technical means of adjusting the model parameters of the initial model through the output matching results and style labels to obtain a feature recognition model, so at least partially overcomes the technical problem of poor generalization of unfamiliar items and styles using machine learning algorithms in related technologies, and improves the The generalization of recognition of unfamiliar objects and unfamiliar styles.

According to an embodiment of the present disclosure, the item image features and text features are input into the image combination layer of the initial model, and the output item combination features include:

The image combination layer is used to stitch item image features and text features into item combination features.

According to the embodiment of the present disclosure, both the image feature and text feature of the item can be represented by a vector, and the image combination layer can be used to stitch the item image feature vector and the text feature vector into a one-dimensional vector, for example: the item image feature vector is (a, b, c), the text feature vector is (e, f, g), then the spliced item combination feature vector can be (a, b, c, e, f, g).

According to the embodiments of the present disclosure, the image combination layer can also be used to stitch the item image feature vector and the text feature vector into a two-dimensional vector or matrix, which will not be repeated here.

According to the embodiments of the present disclosure, the combination feature of the item is obtained by splicing, and the image feature and the text feature of the item can be combined to identify the style matching the item and improve the accuracy of the model prediction.

According to an embodiment of the present disclosure, after acquiring the item sample data set, the above training method further includes:

Generate an augmented item sample dataset from the item sample dataset.

According to an embodiment of the present disclosure, the augmented item sample data set includes on the basis of the acquired item sample data set, and the image data and text data in the item sample data set are respectively processed in two ways: image data enhancement and text data enhancement. Depth representation, the resulting augmented item sample dataset.

According to an embodiment of the present disclosure, for example, the item image data included in the item sample data may be a picture of a black sneaker with a certain brand logo, the text data used to describe the item may be a certain brand of black sneakers, and the item style data may be Sports style, item style tag is 1. The item sample data can be augmented according to the common characteristics of the item sample data. For example, the augmented item sample data can be the item image data of red or green sneakers with a certain brand logo, and the text data used to describe the item. and item style data, item style tags, item image data, text data for describing items, and item image data for sports shoes of various colors that do not carry a certain brand logo and whose styles are similar to those in the item sample data Item style data, style tags, etc. The item sample data augmented by the item sample data set listed above are all positive sample data that match the item and style, so the style label is 1.

According to an embodiment of the present disclosure, when performing model training, negative sample data that does not match the item and style may also be used. For example, the item image data included in the negative sample data may be a picture of black sneakers with a The text data describing the item can be a certain brand of black sneakers, the item style data is business style, and the style tag is 0.

According to the embodiment of the present disclosure, when implementing the feature recognition model training method of the embodiment of the present disclosure, the ratio of the positive sample data to the negative sample data may range from 1:1 to 1:5.

According to an embodiment of the present disclosure, since the item sample data set is mainly derived from domain expert data and a small amount of manually labeled data, the data flow is limited, an augmented item sample data set is generated, and the augmented item sample data set and the item sample data are used to train the initial The model can improve the accuracy of the recognition prediction of the feature recognition model.

The method shown in FIG. 2 will be further described below with reference to FIGS. 3 to 7 in conjunction with specific embodiments.

FIG. 3 schematically shows a flowchart of a method for generating an amplified sample data set according to an embodiment of the present disclosure.

As shown in FIG. 3 , the generating of the amplified sample data set in this embodiment includes operations S310 to S330.

In operation S310, a first item data list similar to the item image data is determined from the item database according to the item image data, wherein the first item data list includes item image data of different items, text data for describing the item, and Item style data.

According to an embodiment of the present disclosure, for example, taking the item image data as a picture of a pink dress with lace as an example, it can be determined from the item database that the first item data list that is similar to the item image data may include a plurality of different items. Items such as: red dress, skirt with lace, pink dress with polka dots. Taking the red dress item in the first item data list as an example, the first item data list includes a picture of the red dress, for example: a picture of a little girl wearing the red dress; text data used to describe the red dress, For example: red, dress; item style data, for example: sweet style, lady style.

In operation S320, a second item data list similar to the text data for describing the item is determined from the item database according to the text data for describing the item, wherein the second item data list includes item image data of different items, Text data and item style data used to describe the item.

According to an embodiment of the present disclosure, for example, taking the text data of a pink dress with lace as an example, from the item database, it can be determined that the second item data list that is similar to the text data used to describe the item includes a plurality of different items. Items such as: skirt with lace, red dress with lace, pink dress with polka dots. Take the second item data list including a skirt with lace as an example, the item image data included in the second item data list is a picture of a skirt with lace, and the text data used to describe the item is lace Lace, skirt, the style data of the item is lady style, sweet style.

In operation S330, an augmented item sample data set is generated according to the first item data list and the second item data list.

According to an embodiment of the present disclosure, an augmented item sample data set can be generated according to all data in the first item data list and the second item data list, and the generated augmented item sample data set includes a red dress with lace Item image data, text data for describing the item, and style data for the item.

According to the embodiments of the present disclosure, the item data list is generated by the item image data and the text data used to describe the item respectively, and the item sample data can be automatically expanded, so as to solve the problem that the related art only relies on the acquired sample data for model training, resulting in The problem of low model prediction accuracy and poor generalization.

According to an embodiment of the present disclosure, generating an augmented item sample data set according to the first item data list and the second item data list, including:

An augmented item sample data set is generated according to the data intersection of the first item data list and the second item data list.

According to an embodiment of the present disclosure, for example, the first item data list includes a red dress, a skirt with lace, and a pink dress with polka dots. The second item data list includes a skirt with lace, a red dress with lace, and a pink dress with polka dots. The generated augmented item sample data set includes the image data of the item as a picture of a pink skirt with lace, and the text data used to describe the item includes: lace, pink, dress, skirt, red, and the style of the item includes : Lady style, sweet style.

According to the embodiment of the present disclosure, an item data list is separately generated from the item image data and the text data used to describe the item, and then the intersection of the two data lists is obtained. While achieving automatic amplification of sample data, it also improves the post-amplification performance. Accuracy of sample data.

According to an embodiment of the present disclosure, the model parameters of the initial model include model parameters of the image feature extraction layer of the initial model, model parameters of the text feature extraction layer of the initial model, and model parameters of the style feature extraction layer of the initial model. According to the matching result and The style label adjusts the model parameters of the initial model to obtain the trained feature recognition model, including:

According to the matching results and style labels, the model parameters of the image feature extraction layer of the initial model, the model parameters of the text feature extraction layer of the initial model, and the model parameters of the style feature extraction layer of the initial model are adjusted to obtain the trained feature recognition model.

According to an embodiment of the present disclosure, the style tag can be represented by a binary number. When the style tag is 1, it means that the item in the item sample data matches the style; when the style tag is 0, it means that the item in the item sample data matches the style. Style does not match.

According to an embodiment of the present disclosure, for example, after inputting the sample data of a genuine item that matches the item and style into the initial model, the output matching result is 0.8, and the style label is 1, then the image feature extraction layer, text can be adjusted by adjusting the initial model. The model parameters of the feature extraction layer and the style feature extraction layer change the output matching result.

According to the embodiment of the present disclosure, the formula shown in formula (1) can be used to calculate the difference between the matching result and the actual result.

Among them, XOR is the exclusive OR function. When the two numbers with the value of 0 or 1 are equal, the XOR function returns a value of 0, otherwise it is 1;

(Item _i , Style _j ) _true is the matching relationship between the actual item and the style, the value is 0 or 1, 1 means corresponding, 0 means not corresponding;

(Item _i , Style _j ) _pred is the matching relationship between the item and the style predicted by the algorithm. The value range is a floating point number between 0 and 1. The larger the value, the higher the matching degree between the item and the style is considered by the algorithm;

I(·) is an indicator function, here it returns 0 when the input value is less than 0, and it returns 1 when the input value is greater than 0.

According to the embodiments of the present disclosure, multiple matching results and style labels output after adjusting the model parameters for many times can be input into the loss function. When the change of the loss function approaches zero, at this time, the image feature extraction layer, The image features, text features and style features extracted by the text feature extraction layer and style feature extraction layer have the highest matching degree, indicating that the prediction accuracy of the feature recognition model is the highest, then the training of the initial model is completed, and the trained feature recognition model is obtained.

According to the embodiments of the present disclosure, the image feature extraction layer, text feature extraction layer, and style feature extraction layer of the initial model are adjusted through the matching result outputted by the matching layer and the style label, so as to realize the supervised model training process and improve the accuracy of model recognition and prediction. Spend.

FIG. 4 schematically shows a schematic diagram of the architecture of a feature recognition model according to an embodiment of the present disclosure.

As shown in FIG. 4 , the feature recognition model of this embodiment includes a three-layer architecture: the bottom layer includes an image feature extraction layer, a text feature extraction layer, and a style feature extraction layer, which are respectively used to extract image features and text features in the item sample data. and style characteristics. The middle layer includes a feature combination layer, which is used for combining image features and text features to form item combination features. The top layer includes a matching layer, which is used to match the item combination features with the style features, output the matching results, and adjust the bottom-most model parameters according to the matching results and style labels.

Fig. 5 schematically shows a flow chart of a method for identifying a feature of an item according to an embodiment of the present disclosure.

As shown in FIG. 5 , the object feature identification method of this embodiment includes operations S510 to S560.

In operation S510, item information of the item to be processed is acquired, wherein the item information includes item image information and text information for describing the item.

According to an embodiment of the present disclosure, for example, the item image information included in the item information with the processed item is a picture of a dark blue denim jacket with fur collar and a certain brand's exclusive pattern, and the text information used to describe the item is a certain brand of dark blue Fur collar denim jacket.

In operation S520, the item image information is input into the image feature extraction layer of the feature recognition model, and the item image feature of the item to be processed is output, wherein the feature recognition model is trained by the training method of the embodiment of the present disclosure.

According to an embodiment of the present disclosure, a picture of a dark blue denim jacket with a fur collar and an exclusive pattern of a certain brand is input into an image feature extraction layer of a feature recognition model obtained by training with the training method provided by the embodiment of the present disclosure, and the output image features of the item are Can include: dark blue, fur collar, denim jacket, a brand-specific pattern, such as a four-leaf clover pattern.

In operation S530, the text information for describing the item is input into the text feature extraction layer of the feature recognition model, and the text feature of the item to be processed is output.

According to the embodiment of the present disclosure, the text information of a certain brand of dark blue fur-collar denim jacket is input into the text feature extraction layer of the feature recognition model, and the output text features of the item to be processed may include: dark blue, a certain brand, fur collar, denim jacket.

In operation S540, the item image feature and the text feature are input into the feature combination layer of the feature recognition model, and the item combination feature of the item to be processed is output.

According to an embodiment of the present disclosure, the object image feature "dark blue, fur collar, denim jacket, a brand-specific pattern, such as a four-leaf clover pattern" is combined with the text feature "dark blue, a certain brand, fur collar, denim jacket" "Input the feature combination layer of the feature recognition model, and output the item combination feature of the item to be processed, which can be "dark blue, fur collar, denim jacket, four-leaf clover pattern, a certain brand".

In operation S550, the item combination feature and the description feature of the candidate style are input into the matching layer of the feature recognition model, and a matching result for characterizing the item combination feature and the style feature is output, wherein the description feature of the candidate style is obtained from the candidate style database. The obtained item style information is obtained by inputting the style feature extraction layer of the feature recognition model.

According to an embodiment of the present disclosure, for example, the item combination feature "dark blue, fur collar, denim jacket, four-leaf clover pattern, a certain brand" and the description feature of the candidate style, for example, the description feature of the candidate style is casual style, enter The matching layer of the feature recognition model outputs the matching results.

In operation S560, according to the matching result, determine the style feature information in which the commodity style information of the item to be processed matches the item combination feature of the item to be processed.

According to the embodiment of the present disclosure, a plurality of style features can be obtained after obtaining a plurality of item style information from the candidate style database and inputting the style feature extraction layer of the feature recognition model. Input the item combination feature and multiple style features into the matching layer of the feature recognition model, and sort the obtained multiple matching results. If only one style is reserved for the item to be processed, the style with the highest matching result can be used as the item style of the item to be processed. information. If the items to be processed can retain N styles, the styles corresponding to the N matching results can be sequentially selected from the sorted matching results from top to bottom as the item style information of the items to be processed.

According to the embodiments of the present disclosure, through the feature recognition model, it is possible to identify the item image features and text features that match the style, and combine the item image features and text features to form the item combination feature, and use the style feature extraction layer in the feature recognition model. The style feature is extracted from the candidate style database, and the style feature matching the combination feature of the item is determined, so as to determine the style of the item, and the technical effect of automatically determining the style of the new item is achieved.

FIG. 6 schematically shows a flowchart of a method for generating a candidate style database according to an embodiment of the present disclosure.

As shown in FIG. 6 , the method for generating a candidate style database in this embodiment includes: S610-S630.

In operation S610, a plurality of item review text data and a plurality of item title text data are acquired.

According to an embodiment of the present disclosure, the article review text data may include that this piece of clothing is suitable for home, travel, or any leisure occasion, is comfortable to wear, and the like. Item title text data may include home, travel essentials, comfort, and the like.

In operation S620, a plurality of item review text data and a plurality of item title text data are preprocessed to obtain a text data set for characterizing the style of the item.

According to an embodiment of the present disclosure, preprocessing may include cleaning the data, such as word segmentation, stop word removal, punctuation removal, feature symbol removal, filtering of high-frequency words, filtering of low-frequency words, and the like. The preprocessing also includes performing a clustering calculation on the cleaned data to obtain a text data set used to characterize the style of the item. For example: leisure, sports, etc.

In operation S630, a candidate style database is generated according to the text data set used to characterize the style of the item.

According to an embodiment of the present disclosure, a text data set used to characterize the style of an item can be stored in a database to generate a candidate style database.

According to an embodiment of the present disclosure, by acquiring the comment data and title data of the item, and through data preprocessing, a text data set for characterizing the style of the item is obtained. It solves the problem of lack of candidate style database or the difficulty of automatically generating candidate style database in machine learning process.

According to an embodiment of the present disclosure, the method for generating a candidate style database further includes:

Generate a vector dataset for characterizing item style from a text dataset for characterizing item style.

Generate a candidate style database from the vector dataset used to characterize item styles.

According to the embodiments of the present disclosure, the data in the candidate style database can be represented by a vector, and when matching styles for new items, the style data can be queried from the candidate style database through vector retrieval, which can save matching time.

In order to compare the prediction time of vector retrieval and ordinary retrieval, the complexity of prediction time of the two retrieval methods can be expressed by formula (2) and formula (3).

cost=|I|*|J|*d (2)

where t is the number of iterations, |J| is the number of styles, k is the number of cluster centers, |I| is the number of items, c is the number of nearest cluster centers to find, and d is the vector dimension of the model training results.

Equation (2) represents the predicted time complexity of ordinary retrieval, and Equation (3) represents the predicted time complexity of violent retrieval.

Because both t and c are constants, and k is a number much smaller than |J| and |I|, the calculation result of formula (3) is much smaller than the calculation result of formula (2), even if the vector retrieval algorithm is used than the predicted time period of common retrieval algorithms.

FIG. 7 schematically shows a system architecture diagram of item feature recognition according to an embodiment of the present disclosure.

As shown in Figure 7, the item feature recognition system includes three parts: a model training module, a candidate style database generation module and a model prediction module.

The model training module augments the sample data by acquiring domain expert data and manual annotation data, and uses the augmented sample data to train the initial model to obtain a feature recognition model.

The candidate style database generation module obtains the candidate style database according to the method for generating the candidate style database in the embodiment of the present disclosure by acquiring the item review data and the item title data.

The model prediction module inputs the item image data of the item to be processed into the image feature extraction layer to extract the image feature of the item to be processed; extracts the text feature of the item to be processed from the text data of the item to be processed which is used to describe the item; and uses the item feature combination layer The image feature and text feature of the item to be processed are combined to obtain the item combination feature. Match the item combination feature with the description feature of the candidate style extracted by the style feature extraction layer in the matching layer to determine the style of the item to be processed.

FIG. 8 schematically shows a block diagram of an apparatus for training a feature recognition model according to an embodiment of the present disclosure.

As shown in FIG. 8 , the feature recognition model training apparatus 800 includes: a first acquisition module 810 , a feature extraction module 820 , a feature combination module 830 , a matching module 840 and an adjustment module 850 .

The first acquisition module 810 is configured to acquire an item sample data set, wherein the item sample data set includes a plurality of item samples, and each item sample includes item image data, text data for describing the item, and item style data. Samples have object style labels.

The feature extraction module 820 is used for, for each item sample, input the item image data into the image feature extraction layer of the initial model, and output the item image features; input the text data used to describe the item into the text feature extraction layer of the initial model, and output the text Features; input item style data into the style feature extraction layer of the initial model, and output style features.

The feature combination module 830 is used for inputting the item image features and text features into the feature combination layer of the initial model, and outputting the item combination features.

The matching module 840 is configured to input the item combination feature and the style feature into the matching layer of the initial model, and output a matching result for characterizing the item combination feature and the style feature.

The adjustment module 850 is configured to adjust the model parameters of the initial model according to the matching result and the style label, so as to obtain a trained feature recognition model.

According to an embodiment of the present disclosure, the feature combination module 830 includes a stitching unit for stitching the item image feature and the text feature into the item combination feature using the image combination layer.

According to an embodiment of the present disclosure, the above-mentioned apparatus further includes: a generating module configured to generate an augmented item sample data set according to the item sample data set.

According to an embodiment of the present disclosure, the generating module includes a first determining unit, a second determining unit, and a generating unit. The first determining unit determines a first item data list similar to the item image data from the item database according to the item image data, wherein the first item data list includes item image data of different items, a description of the item for describing the item Text data and item style data. The second determining unit is configured to determine a second item data list similar to the text data for describing the item from the item database according to the text data for describing the item, wherein the second item data list includes items of different items Image data, text data to describe the item, and item style data. The generating unit is configured to generate an augmented item sample data set according to the first item data list and the second item data list.

According to an embodiment of the present disclosure, the generating unit includes a generating subunit for generating an augmented item sample data set according to the data intersection of the first item data list and the second item data list.

According to an embodiment of the present disclosure, the adjustment module includes an adjustment unit for adjusting the model parameters of the image feature extraction layer of the initial model, the model parameters of the text feature extraction layer of the initial model, and the style features of the initial model according to the matching result and the style label Extract the model parameters of the layer to obtain the trained feature recognition model.

FIG. 9 schematically shows a block diagram of an apparatus for identifying characteristics of an item according to an embodiment of the present disclosure.

As shown in FIG. 9 , the article feature identification device of this embodiment includes a second acquisition module 910 , a first identification module 920 , a second identification module 930 , a third identification module 940 , a fourth identification module 950 and a determination module 960 .

The second acquiring module 910 is configured to acquire item information of the item to be processed, wherein the item information includes item image information and text information for describing the item.

The first recognition module 920 is configured to input the image information of the item into the image feature extraction layer of the feature recognition model, and output the item image features of the item to be processed, wherein the feature recognition model is trained by the training method provided by the embodiment of the present disclosure.

The second recognition module 930 is configured to input the text information for describing the item into the text feature extraction layer of the feature recognition model, and output the text feature of the item to be processed.

The third recognition module 940 is configured to input the image feature and the text feature of the item into the feature combination layer of the feature recognition model, and output the item combination feature of the item to be processed.

The fourth identification module 950 is configured to input the item combination feature and the description feature of the candidate style into the matching layer of the feature recognition model, and output a matching result used to characterize the item combination feature and the description feature of the candidate style , wherein the description feature of the candidate style is obtained by inputting the item style information obtained from the candidate style database into the style feature extraction layer of the feature recognition model.

The determining module 960 is configured to determine, according to the matching result, style feature information matching the item combination feature of the item to be processed. It should be noted that, the embodiments of the apparatus part of the present disclosure correspond to or are similar to the embodiments of the method part of the present disclosure, and details are not described herein again.

Any of the modules, units, sub-units, or at least part of the functions of any of the modules, units, and sub-units according to the embodiments of the present disclosure may be implemented in one module. Any one or more of the modules, units, and sub-units according to the embodiments of the present disclosure may be divided into multiple modules for implementation. Any one or more of the modules, units, sub-units according to embodiments of the present disclosure may be implemented at least partially as hardware circuits, such as field programmable gate arrays (FPGAs), programmable logic arrays (PLAs), systems on chips, A system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or any other reasonable hardware or firmware implementation that can integrate or package a circuit, or in software, hardware, and firmware. Any one of them or an appropriate combination of any of them can be implemented. Alternatively, one or more of the modules, units, and sub-units according to the embodiments of the present disclosure may be implemented at least in part as computer program modules, which, when executed, may perform corresponding functions.

For example, the first acquisition module 810, the feature extraction module 820, the feature combination module 830, the matching module 840, the adjustment module 850 or the second acquisition module 910, the first identification module 920, the second identification module 930, the third identification module 940, Any number of the fourth identification module 950 and the determination module 960 may be combined into one module/unit/subunit, or any one of the modules/units/subunits may be split into multiple modules/units/subunits. unit. Alternatively, at least part of the functionality of one or more of these modules/units/subunits may be combined with at least part of the functionality of other modules/units/subunits and combined in one module/unit/subunit realized in. According to an embodiment of the present disclosure, the first acquisition module 810 , the feature extraction module 820 , the feature combination module 830 , the matching module 840 , the adjustment module 850 or the second acquisition module 910 , the first identification module 920 , the second identification module 930 , the first At least one of the three identification module 940, the fourth identification module 950, and the determination module 960 may be implemented at least in part as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a substrate A system on a package, a system on a package, an application specific integrated circuit (ASIC), or any other reasonable way to integrate or package a circuit, such as hardware or firmware, or in software, hardware, and firmware. Any one or an appropriate combination of any of them is implemented. Or, the first acquisition module 810, the feature extraction module 820, the feature combination module 830, the matching module 840, the adjustment module 850 or the second acquisition module 910, the first recognition module 920, the second recognition module 930, the third recognition module 940, At least one of the fourth identification module 950 and the determination module 960 may be implemented at least in part as a computer program module that, when executed, may perform corresponding functions.

Figure 10 schematically shows a block diagram of an electronic device suitable for implementing the method described above, according to an embodiment of the present disclosure. The electronic device shown in FIG. 10 is only an example, and should not impose any limitation on the function and scope of use of the embodiments of the present disclosure.

As shown in FIG. 10 , an electronic device 1000 according to an embodiment of the present disclosure includes a processor 1001 that can be loaded into a random access memory (RAM) 1003 according to a program stored in a read only memory (ROM) 1002 or from a storage part 1008 program to perform various appropriate actions and processes. The processor 1001 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and/or a related chipset, and/or a special-purpose microprocessor (e.g., an application specific integrated circuit (ASIC)), among others. The processor 1001 may also include onboard memory for caching purposes. The processor 1001 may include a single processing unit or multiple processing units for performing different actions of the method flow according to the embodiments of the present disclosure.

In the RAM 1003, various programs and data necessary for the operation of the electronic device 1000 are stored. The processor 1001 , the ROM 1002 and the RAM 1003 are connected to each other through a bus 1004 . The processor 1001 performs various operations of the method flow according to the embodiment of the present disclosure by executing the programs in the ROM 1002 and/or the RAM 1003 . Note that the program may also be stored in one or more memories other than the ROM 1002 and the RAM 1003 . The processor 1001 may also perform various operations of the method flow according to the embodiments of the present disclosure by executing programs stored in the one or more memories.

According to an embodiment of the present disclosure, the electronic device 1000 may also include an input/output (I/O) interface 1005 that is also connected to the bus 1004 . System 1000 may also include one or more of the following components connected to I/O interface 1005: input portion 1006 including keyboard, mouse, etc.; including components such as cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers An output section 1007 including a hard disk, etc.; a storage section 1008 including a hard disk, etc.; and a communication section 1009 including a network interface card such as a LAN card, a modem, and the like. The communication section 1009 performs communication processing via a network such as the Internet. A drive 1010 is also connected to the I/O interface 1005 as needed. A removable medium 1011, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is mounted on the drive 1010 as needed so that a computer program read therefrom is installed into the storage section 1008 as needed.

According to an embodiment of the present disclosure, the method flow according to an embodiment of the present disclosure may be implemented as a computer software program. For example, embodiments of the present disclosure include a computer program product comprising a computer program carried on a computer-readable storage medium, the computer program containing program code for performing the method illustrated in the flowchart. In such an embodiment, the computer program may be downloaded and installed from the network via the communication portion 1009, and/or installed from the removable medium 1011. When the computer program is executed by the processor 1001, the above-described functions defined in the system of the embodiment of the present disclosure are executed. According to embodiments of the present disclosure, the above-described systems, apparatuses, apparatuses, modules, units, etc. can be implemented by computer program modules.

The present disclosure also provides a computer-readable storage medium. The computer-readable storage medium may be included in the device/apparatus/system described in the above embodiments; it may also exist alone without being assembled into the device/system. device/system. The above-mentioned computer-readable storage medium carries one or more programs, and when the above-mentioned one or more programs are executed, implement the method according to the embodiment of the present disclosure.

According to an embodiment of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium. Examples may include, but are not limited to, portable computer disks, hard disks, random access memory (RAM), read only memory (ROM), erasable programmable read only memory (EPROM or flash memory), portable compact disk read only memory (CD- ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this disclosure, a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

For example, according to embodiments of the present disclosure, a computer-readable storage medium may include one or more memories other than ROM 1002 and/or RAM 1003 and/or ROM 1002 and RAM 1003 described above.

The flowchart and block diagrams in the Figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code that contains one or more logical functions for implementing the specified functions executable instructions. It should also be noted that, in some alternative implementations, the functions noted in the blocks 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 is also noted that each block of the block diagrams or flowchart illustrations, and combinations of blocks in the block diagrams or flowchart illustrations, can be implemented in special purpose hardware-based systems that perform the specified functions or operations, or can be implemented using A combination of dedicated hardware and computer instructions is implemented.

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

Embodiments of the present disclosure have been described above. However, these examples are for illustrative purposes only, and are not intended to limit the scope of the present disclosure. Although the various embodiments are described above separately, this does not mean that the measures in the various embodiments cannot be used in combination to advantage. The scope of the present disclosure is defined by the appended claims and their equivalents. Without departing from the scope of the present disclosure, those skilled in the art can make various substitutions and modifications, and these substitutions and modifications should all fall within the scope of the present disclosure.

Claims (14)

1. A feature recognition model training method comprises the following steps:

acquiring an article sample data set, wherein the article sample data set comprises a plurality of article samples, each article sample comprises article image data, text data for describing an article and article style data, and the article samples are provided with style labels;

inputting the article image data into an image feature extraction layer of an initial model and outputting article image features aiming at each article sample; inputting the text data for describing the article into a text feature extraction layer of the initial model, and outputting text features; inputting the article style data into a style feature extraction layer of the initial model, and outputting style features;

inputting the article image features and the text features into a feature combination layer of the initial model, and outputting article combination features;

inputting the article combination characteristics and the style characteristics into a matching layer of the initial model, and outputting a matching result for representing the article combination characteristics and the style characteristics;

and adjusting the model parameters of the initial model according to the matching result and the style label to obtain a trained feature recognition model.

2. The training method of claim 1, wherein the item image features and the text features are input into an image combination layer of the initial model, and outputting item combination features comprises:

and splicing the article image features and the text features into the article combination features by utilizing the image combination layer.

3. The training method of claim 1, wherein after acquiring the sample set of article data, further comprising:

and generating an amplified article sample data set according to the article sample data set.

4. The training method of claim 3, wherein said generating an augmented article sample data set from said article sample data set comprises:

determining a first item data list similar to the item image data from an item database according to the item image data, wherein the first item data list comprises item image data of different items, text data for describing the items and item style data;

according to the text data for describing the article, determining a second article data list similar to the text data for describing the article from the article database, wherein the second article data list comprises article image data of different articles, text data for describing the article and article style data;

and generating the amplified article sample data set according to the first article data list and the second article data list.

5. The training method of claim 4, wherein said generating the augmented item sample dataset from the first item data list and the second item data list comprises:

and generating the amplified article sample data set according to the data intersection of the first article data list and the second article data list.

6. The training method according to claim 1, wherein the model parameters of the initial model include model parameters of an image feature extraction layer of the initial model, model parameters of a text feature extraction layer of the initial model, and model parameters of a style feature extraction layer of the initial model, and the adjusting the model parameters of the initial model according to the matching result and the style label to obtain the trained feature recognition model includes:

and adjusting the model parameters of the image feature extraction layer of the initial model, the model parameters of the text feature extraction layer of the initial model and the model parameters of the style feature extraction layer of the initial model according to the error between the matching result and the style label to obtain the trained feature recognition model.

7. An article feature identification feature recognition method comprises the following steps:

acquiring article information of an article to be processed, wherein the article information comprises article image information and text information for describing the article;

inputting the article image information into an image feature extraction layer of a feature recognition model, and outputting article image features of the article to be processed, wherein the feature recognition model is obtained by training through the training method of any one of claims 1 to 6;

inputting the text information for describing the article into a text feature extraction layer of the feature recognition model, and outputting the text feature of the article to be processed;

inputting the article image features and the text features into a feature combination layer of the feature recognition model, and outputting article combination features of the article to be processed;

inputting the item combination characteristics and the description characteristics of the candidate style into a matching layer of the characteristic identification model, and outputting a matching result for representing the item combination characteristics and the description characteristics of the candidate style, wherein the description characteristics of the candidate style are obtained by inputting item style information acquired from a candidate style database into a style characteristic extraction layer of the characteristic identification model;

and determining style characteristic information matched with the article combination characteristics of the article to be processed according to the matching result.

8. The method of claim 7, further comprising:

acquiring a plurality of item comment text data and a plurality of item title text data;

preprocessing the plurality of item comment text data and the plurality of item title text data to obtain a text data set for representing the style of the item;

and generating the candidate style database according to the text data set for representing the style of the article.

9. The method of claim 8, further comprising:

generating a vector data set for representing the style of the article according to the text data set for representing the style of the article;

and generating the candidate style database according to the vector data set for characterizing the style of the article.

10. A feature recognition model training apparatus comprising:

the system comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the article sample data set comprises a plurality of article samples, each article sample comprises article image data, text data for describing an article and article style data, and the article sample has a style label;

the feature extraction module is used for inputting the article image data into an image feature extraction layer of an initial model and outputting article image features aiming at each article sample; inputting the text data for describing the article into a text feature extraction layer of the initial model, and outputting text features; inputting the article style data into a style feature extraction layer of the initial model, and outputting style features;

the feature combination module is used for inputting the article image features and the text features into a feature combination layer of the initial model and outputting article combination features;

the matching module is used for inputting the article combination characteristics and the style characteristics into a matching layer of the initial model and outputting a matching result for representing the article combination characteristics and the style characteristics;

and the adjusting module is used for adjusting the model parameters of the initial model according to the matching result and the style label to obtain a trained feature recognition model.

11. An article feature identification device comprising:

the second acquisition module is used for acquiring article information of the article to be processed, wherein the article information comprises article image information and text information used for describing the article;

a first recognition module, configured to input the article image information into an image feature extraction layer of a feature recognition model, and output an article image feature of the article to be processed, where the feature recognition model is obtained by training according to the training method of any one of claims 1 to 6;

the second identification module is used for inputting the text information for describing the article into a text feature extraction layer of the feature identification model and outputting the text feature of the article to be processed;

the third identification module is used for inputting the article image characteristics and the text characteristics into a characteristic combination layer of the characteristic identification model and outputting article combination characteristics of the article to be processed;

a fourth identification module, configured to input the item combination feature and the description feature of the candidate style into a matching layer of the feature identification model, and output a matching result for characterizing the item combination feature and the description feature of the candidate style, where the description feature of the candidate style is obtained by inputting item style information acquired from a candidate style database into a style feature extraction layer of the feature identification model;

and the determining module is used for determining style characteristic information matched with the article combination characteristic of the article to be processed according to the matching result.

12. An electronic device, comprising:

one or more processors;

a storage device for storing one or more programs,

wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to perform the method of any of claims 1-6 or 7-9.

13. A computer readable storage medium having stored thereon executable instructions which, when executed by a processor, cause the processor to perform the method of any one of claims 1 to 6 or 7 to 9.

14. A computer program product comprising a computer program which, when executed by a processor, implements a method according to any one of claims 1 to 6 or 7 to 9.

Images