Method, system and vr terminal for recommending personalized clothes through vr terminal

Abstract

The embodiment of the present disclosure provides a method, a system and a VR terminal for recommending personalized clothes through a VR terminal, first clothes preference feature distribution corresponding to initiative clothes interaction data in a target VR interaction activity is acquired; second clothes preference feature distribution corresponding to passive clothes interaction data of the target VR interaction activity is acquired; clothes preference knowledge graph corresponding to the target VR interaction activity is determined based on the first clothes preference feature distribution and the second clothes preference feature distribution; clothes personalized push entity corresponding to the target VR interaction activity is acquired based on the clothes preference knowledge graph, so as to recommend clothes for the VR interaction activity.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of VR, in particular, to a method and system for recommending personalized clothes through a VR terminal and the VR terminal. BACKGROUND

[0002] Virtual reality technology includes computer, electronic information, simulation technology, and its basic implementation is mainly based on computer technology, which utilizes and integrates three-dimensional graphics technology, multimedia technology, simulation technology, display technology, servo technology and other latest developments of various high-techs, and produces a virtual world with realistic three-dimensional vision, touch, smell and other sensory experiences through computers and other devices, so that people in the virtual world have a sense of being there. With the continuous development of social productivity and science and technology, the demand for VR technology in various industries is increasing. VR technology has made great progress and has gradually become a new scientific and technological field. For example, in the field of clothing recommendation, how to effectively recommend clothes based on a large number of VR interactive activities is a major technical problem currently studied in the field. SUMMARY

[0003] Therefore, the present disclosure provides a method for recommending personalized clothes through a VR terminal, which is characterized by comprising:

[0004] obtaining a first clothing preference feature distribution corresponding to active clothing interaction data in a target VR interactive activity;

[0005] obtaining a second clothing preference feature distribution corresponding to passive clothing interaction data in the target VR interactive activity;

[0006] determining a clothing preference knowledge graph corresponding to the target VR interactive activity based on the first clothing preference feature distribution and the second clothing preference feature distribution;

[0007] obtaining a clothing personalized push entity corresponding to the target VR interactive activity based on the clothing preference knowledge graph, so as to recommend clothes for the VR interactive activity.

[0008] In some examples, the determination of the clothing preference knowledge graph corresponding to the target VR interactive activity based on the first clothing preference feature distribution and the second clothing preference feature distribution comprises:

[0009] perform attention feature extraction on the first clothing preference feature distribution and the second clothing preference feature distribution based on the clothing preference heat map corresponding to the first clothing preference feature distribution and the clothing preference heat map corresponding to the second clothing preference feature distribution, to obtain N attention features corresponding to the target VR interaction activity; N is a sum of a category number of the first clothing preference feature distribution and a category number of the second clothing preference feature distribution;

[0010] select M attention features from the N attention features as a clothing preference knowledge graph corresponding to the target VR interaction activity based on clothing preference heat maps corresponding to the N attention features.

[0011] In some examples, further comprising:

[0012] obtaining a target VR interaction activity and extracting initiative clothing interaction data in the target VR interaction activity;

[0013] performing preference feature decision on the initiative clothing interaction data by a preference feature decision model to obtain an initial clothing preference feature distribution corresponding to the initiative clothing interaction data and a clothing preference heat map corresponding to the initial clothing preference feature distribution;

[0014] obtaining an associated clothing preference feature distribution corresponding to the initial clothing preference feature distribution;

[0015] fusing the initial clothing preference feature distribution and the associated clothing preference feature distribution, and determining the fused initial clothing preference feature distribution as a first clothing preference feature distribution corresponding to the initiative clothing interaction data;

[0016] determining the clothing preference heat map corresponding to the initial clothing preference feature distribution as a clothing preference heat map corresponding to the first clothing preference feature distribution.

[0017] In some examples, the performing of the preference feature decision on the initiative clothing interaction data by the preference feature decision model to obtain the initial clothing preference feature distribution corresponding to the initiative clothing interaction data and the clothing preference heat map corresponding to the initial clothing preference feature distribution comprises:

[0018] extracting, by the preference feature decision model, interaction key feedback data corresponding to the initiative clothing interaction data;

[0019] inputting the interaction key feedback data into a decision network layer in the preference feature decision model to obtain a correlation parameter between the interaction key feedback data and target clothing preference feature distribution information in the decision network layer;

[0020] determine a garment preference feature distribution corresponding to the target garment preference feature distribution information with the largest correlation parameter as an initial garment preference feature distribution corresponding to the active garment interaction data;

[0021] calculate a garment preference heat map corresponding to the initial garment preference feature distribution based on the largest correlation parameter corresponding to the target garment preference feature distribution information.

[0022] In some examples, further comprising:

[0023] extract an interaction data interval of a VR interaction node and an interaction mapping partition of the VR interaction node from the target VR interaction activity;

[0024] obtain a relevant VR display knowledge point corresponding to the interaction data interval of the VR interaction node, and in the relevant VR display knowledge point, obtain a first node garment preference feature distribution corresponding to the interaction data interval of the VR interaction node as a first associated garment preference feature distribution, and configure a first garment preference heat map for the first associated garment preference feature distribution;

[0025] obtain a relevant VR display knowledge point corresponding to the interaction mapping partition of the VR interaction node, and in the relevant VR display knowledge point, obtain a second node garment preference feature distribution corresponding to the interaction mapping partition of the VR interaction node as a second associated garment preference feature distribution, and configure a second garment preference heat map for the second associated garment preference feature distribution;

[0026] determine passive garment interaction data of the target VR interaction activity based on the first associated garment preference feature distribution and the second associated garment preference feature distribution, and obtain a second garment preference feature distribution corresponding to the passive garment interaction data;

[0027] determine a garment preference heat map corresponding to the second garment preference feature distribution based on the passive garment interaction data, the first garment preference heat map, and the second garment preference heat map.

[0028] The present disclosure also provides a system for recommending personalized garments through a VR terminal, comprising:

[0029] a first obtaining unit configured to obtain a first garment preference feature distribution corresponding to active garment interaction data in a target VR interaction activity;

[0030] a second obtaining unit configured to obtain a second garment preference feature distribution corresponding to passive garment interaction data of the target VR interaction activity;

[0031] determine, based on the first clothing preference feature distribution and the second clothing preference feature distribution, a clothing preference knowledge graph corresponding to the target VR interactive activity;

[0032] a clothing recommendation unit configured to obtain a clothing personalized push entity corresponding to the target VR interactive activity based on the clothing preference knowledge graph, so as to perform clothing recommendation for the VR interactive activity.

[0033] In some examples, the determination unit is specifically configured to:

[0034] perform attention feature extraction on the first clothing preference feature distribution and the second clothing preference feature distribution based on a clothing preference heat map corresponding to the first clothing preference feature distribution and a clothing preference heat map corresponding to the second clothing preference feature distribution, to obtain N attention features corresponding to the target VR interactive activity; N is a sum of a category number of the first clothing preference feature distribution and a category number of the second clothing preference feature distribution.

[0035] select M attention features from the N attention features as the clothing preference knowledge graph corresponding to the target VR interactive activity based on clothing preference heat maps corresponding to the N attention features.

[0036] In some examples, the determination unit is further configured to:

[0037] obtain a target VR interactive activity and extract initiative clothing interaction data in the target VR interactive activity;

[0038] perform preference feature decision on the initiative clothing interaction data through a preference feature decision model, to obtain an initial clothing preference feature distribution corresponding to the initiative clothing interaction data and a clothing preference heat map corresponding to the initial clothing preference feature distribution;

[0039] obtain an associated clothing preference feature distribution corresponding to the initial clothing preference feature distribution;

[0040] fuse the initial clothing preference feature distribution and the associated clothing preference feature distribution, and determine a fused initial clothing preference feature distribution as a first clothing preference feature distribution corresponding to the initiative clothing interaction data;

[0041] determine a clothing preference heat map corresponding to the initial clothing preference feature distribution as a clothing preference heat map corresponding to the first clothing preference feature distribution;

[0042] The preference feature decision model is used to determine the active clothing interaction data, to obtain an initial clothing preference feature distribution corresponding to the active clothing interaction data and a clothing preference heat map corresponding to the initial clothing preference feature distribution, including:

[0043] The preference feature decision model is used to determine the active clothing interaction data, to obtain an initial clothing preference feature distribution corresponding to the active clothing interaction data and a clothing preference heat map corresponding to the initial clothing preference feature distribution, including:

[0044] The interaction key feedback data corresponding to the active clothing interaction data is input into a decision network layer in the preference feature decision model, to obtain a correlation parameter between the interaction key feedback data and target clothing preference feature distribution information in the decision network layer;

[0045] The clothing preference feature distribution corresponding to the target clothing preference feature distribution information with the maximum correlation parameter is determined as the initial clothing preference feature distribution corresponding to the active clothing interaction data;

[0046] The clothing preference heat map corresponding to the initial clothing preference feature distribution is calculated based on the maximum correlation parameter corresponding to the target clothing preference feature distribution information.

[0047] In some examples, the determination unit is further configured to:

[0048] extract an interaction data interval of a VR interaction node and an interaction mapping partition of the VR interaction node from the target VR interaction activity;

[0049] obtain a related VR display knowledge point corresponding to the interaction data interval of the VR interaction node, and in the related VR display knowledge point, a first node clothing preference feature distribution corresponding to the interaction data interval of the VR interaction node is queried as a first associated clothing preference feature distribution, and a first clothing preference heat map is configured for the first associated clothing preference feature distribution;

[0050] obtain a related VR display knowledge point corresponding to the interaction mapping partition of the VR interaction node, and in the related VR display knowledge point, a second node clothing preference feature distribution corresponding to the interaction mapping partition of the VR interaction node is queried as a second associated clothing preference feature distribution, and a second clothing preference heat map is configured for the second associated clothing preference feature distribution;

[0051] determine passive clothing interaction data of the target VR interaction activity based on the first associated clothing preference feature distribution and the second associated clothing preference feature distribution, to obtain a second clothing preference feature distribution corresponding to the passive clothing interaction data;

[0052] Determine a clothing preference heat map corresponding to the second clothing preference feature distribution based on the passive clothing interaction data, the first clothing preference heat map, and the second clothing preference heat map.

[0053] The present disclosure also provides a VR terminal including a processor and a machine-readable storage medium storing machine-executable instructions, and the processor implements the method described above when executing the machine-executable instructions.

[0054] In summary, the present disclosure provides a method, system, and VR terminal for recommending personalized clothing through a VR terminal. The first clothing preference feature distribution corresponding to the active clothing interaction data in the target VR interactive activity can be obtained by intelligently identifying the active clothing interaction data in the target VR interactive activity. The second clothing preference feature distribution corresponding to the passive clothing interaction data of the target VR interactive activity is obtained. The clothing preference knowledge graph corresponding to the target VR interactive activity is determined based on the first clothing preference feature distribution and the second clothing preference feature distribution. The clothing personalized push entity corresponding to the target VR interactive activity can be obtained through the clothing preference knowledge graph to recommend clothing for the VR interactive activity. In this way, the clothing preference knowledge graph describing the target VR interactive activity is obtained, and the clothing personalized push entity corresponding to the target VR interactive activity is quickly searched based on the clothing preference knowledge graph, thereby recommending clothing for the target VR interactive activity, which can improve the reliability of clothing recommendation. BRIEF DESCRIPTION OF DRAWINGS

[0055] To more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings needed in the embodiments will be briefly introduced as follows.

[0056] Figure 1 The flowchart of the method for recommending personalized clothing through a VR terminal provided by the embodiments of the present disclosure is shown.

[0057] Figure 2 The architecture diagram of the VR terminal provided by the embodiments of the present disclosure is shown.

[0058] Figure 3 The functional unit diagram of the system for recommending personalized clothing through a VR terminal provided by the embodiments of the present disclosure is shown. DETAILED DESCRIPTION

[0059] To make the purposes, technical solutions, and advantages of the embodiments of the present disclosure clearer, the technical solutions in the embodiments of the present disclosure will be described clearly and completely below with reference to the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are some but not all of the embodiments of the present disclosure. The components of the embodiments of the present disclosure described and shown in the drawings herein can be arranged and designed in various different configurations.

[0060] As shown in Figure 1 Figure 1 A flowchart of a method for recommending personalized clothes through a VR terminal provided by the present embodiment is shown in the following, and each step of the method will be described in detail.

[0061] Step 10, obtaining a first clothing preference feature distribution corresponding to the initiative clothing interaction data in the target VR interactive activity;

[0062] Step 20, obtaining a second clothing preference feature distribution corresponding to the passive clothing interaction data of the target VR interactive activity;

[0063] Step 30, determining a clothing preference knowledge graph corresponding to the target VR interactive activity based on the first clothing preference feature distribution and the second clothing preference feature distribution;

[0064] Step 40, obtaining a clothing personalized push entity corresponding to the target VR interactive activity based on the clothing preference knowledge graph, to recommend clothes for the VR interactive activity.

[0065] In detail, in the embodiments of the present disclosure, alternatively, step 30 comprises:

[0066] Based on the clothing preference heat map corresponding to the first clothing preference feature distribution, the clothing preference heat map corresponding to the second clothing preference feature distribution, attention feature extraction is performed on the first clothing preference feature distribution and the second clothing preference feature distribution, to obtain N attention features corresponding to the target VR interactive activity; N is the sum of the number of categories of the first clothing preference feature distribution and the number of categories of the second clothing preference feature distribution;

[0067] Based on the clothing preference heat map corresponding to the N attention features, M attention features are selected from the N attention features as the clothing preference knowledge graph corresponding to the target VR interactive activity.

[0068] In detail, in the embodiments of the present disclosure, alternatively, the method further comprises:

[0069] Obtaining a target VR interactive activity, and extracting initiative clothing interaction data in the target VR interactive activity;

[0070] ​The preference feature decision model is used to make a preference feature decision on the active clothing interaction data, to obtain an initial clothing preference feature distribution corresponding to the active clothing interaction data and a clothing preference heat map corresponding to the initial clothing preference feature distribution;

[0071] An associated clothing preference feature distribution corresponding to the initial clothing preference feature distribution is obtained.

[0072] The initial clothing preference feature distribution and the associated clothing preference feature distribution are fused, and the fused initial clothing preference feature distribution is determined as a first clothing preference feature distribution corresponding to the active clothing interaction data.

[0073] The clothing preference heat map corresponding to the initial clothing preference feature distribution is determined as a clothing preference heat map corresponding to the first clothing preference feature distribution.

[0074] In detail, in the embodiment of the present disclosure, alternatively, the preference feature decision model is used to make a preference feature decision on the active clothing interaction data, to obtain an initial clothing preference feature distribution corresponding to the active clothing interaction data and a clothing preference heat map corresponding to the initial clothing preference feature distribution, which includes:

[0075] The interaction key feedback data corresponding to the active clothing interaction data is extracted by the preference feature decision model.

[0076] The interaction key feedback data is input into a decision network layer in the preference feature decision model, to obtain a correlation parameter between the interaction key feedback data and target clothing preference feature distribution information in the decision network layer.

[0077] The clothing preference feature distribution corresponding to the target clothing preference feature distribution information with the maximum correlation parameter is determined as the initial clothing preference feature distribution corresponding to the active clothing interaction data.

[0078] The clothing preference heat map corresponding to the initial clothing preference feature distribution is calculated based on the maximum correlation parameter corresponding to the target clothing preference feature distribution information.

[0079] In detail, in the embodiment of the present disclosure, alternatively, the method further includes:

[0080] An interaction data interval of a VR interaction node and an interaction mapping partition of the VR interaction node are extracted from the target VR interaction activity.

[0081] acquire a relevant VR display knowledge point corresponding to the interaction data interval of the VR interaction node, and in the relevant VR display knowledge point, acquire a first node clothing preference feature distribution corresponding to the interaction data interval of the VR interaction node as a first associated clothing preference feature distribution, and configure a first clothing preference heat map for the first associated clothing preference feature distribution;

[0082] acquire a relevant VR display knowledge point corresponding to the interaction mapping partition of the VR interaction node, and in the relevant VR display knowledge point, acquire a second node clothing preference feature distribution corresponding to the interaction mapping partition of the VR interaction node as a second associated clothing preference feature distribution, and configure a second clothing preference heat map for the second associated clothing preference feature distribution;

[0083] determine passive clothing interaction data of the target VR interaction activity based on the first associated clothing preference feature distribution and the second associated clothing preference feature distribution, and obtain a second clothing preference feature distribution corresponding to the passive clothing interaction data;

[0084] determine a clothing preference heat map corresponding to the second clothing preference feature distribution based on the passive clothing interaction data, the first clothing preference heat map, and the second clothing preference heat map. Figure 2 , Figure 2 The architecture of a VR terminal 20 is provided in this embodiment. The VR terminal 20 can include a processor 21 and a machine-readable storage medium 22. The processor 21 and the machine-readable storage medium 22 can be communicatively connected through a bus. The machine-readable storage medium 22 stores machine-executable instructions. By reading and executing the corresponding machine-executable instructions in the machine-readable storage medium 22, the processor 21 can implement the method of recommending personalized clothing through a VR terminal described above.

[0085] The machine-readable storage medium 22 mentioned herein can be any one or a combination of a radom access memory (RAM), a volatile memory, a non-volatile memory, a flash memory, a storage drive (such as a hard disk drive), and a solid state disk.

[0086] As shown in Figure 3 The embodiment further provides a system 30 for recommending personalized clothing through a VR terminal. The system 30 for recommending personalized clothing through a VR terminal includes a plurality of functional units, specifically a first acquisition unit 301, a second acquisition unit 302, a determination unit 303, and a clothing recommendation unit 304. The system 30 can be the VR terminal described above, or can be a software unit or an integrated circuit unit in the VR terminal, or can be the above-mentioned to-be-processed, and the specific implementation is not limited,

[0087] The first obtaining unit 301 is configured to obtain a first clothing preference feature distribution corresponding to active clothing interaction data in a target VR interaction activity.

[0088] In this embodiment, the first obtaining unit 301 can be configured to perform Figure 1 The specific description of the first obtaining unit 301 can refer to the description of step 10.

[0089] The second obtaining unit 302 is configured to obtain a second clothing preference feature distribution corresponding to passive clothing interaction data in the target VR interaction activity.

[0090] The determining unit 303 is configured to determine a clothing preference knowledge graph corresponding to the target VR interaction activity based on the first clothing preference feature distribution and the second clothing preference feature distribution.

[0091] The clothing recommendation unit 304 is configured to obtain a clothing personalized push entity corresponding to the target VR interaction activity based on the clothing preference knowledge graph, so as to perform clothing recommendation for the VR interaction activity.

[0092] In some examples, the determining unit is specifically configured to:

[0093] perform attention feature extraction on the first clothing preference feature distribution and the second clothing preference feature distribution based on a clothing preference heat map corresponding to the first clothing preference feature distribution and a clothing preference heat map corresponding to the second clothing preference feature distribution, to obtain N attention features corresponding to the target VR interaction activity; N is a sum of a category number of the first clothing preference feature distribution and a category number of the second clothing preference feature distribution.

[0094] select M attention features from the N attention features as the clothing preference knowledge graph corresponding to the target VR interaction activity based on clothing preference heat maps corresponding to the N attention features.

[0095] In some examples, the determining unit is further configured to:

[0096] obtain a target VR interaction activity, and extract active clothing interaction data in the target VR interaction activity.

[0097] perform preference feature decision on the active clothing interaction data through a preference feature decision model, to obtain an initial clothing preference feature distribution corresponding to the active clothing interaction data and a clothing preference heat map corresponding to the initial clothing preference feature distribution.

[0098] obtain an associated clothing preference feature distribution corresponding to the initial clothing preference feature distribution.

[0099] fuse the initial garment preference feature distribution and the associated garment preference feature distribution, and determine a fused initial garment preference feature distribution as the first garment preference feature distribution corresponding to the initiative garment interaction data;

[0100] determine a garment preference heat map corresponding to the initial garment preference feature distribution as a garment preference heat map corresponding to the first garment preference feature distribution;

[0101] The preference feature decision model is used to determine the initiative garment interaction data, and an initial garment preference feature distribution corresponding to the initiative garment interaction data and a garment preference heat map corresponding to the initial garment preference feature distribution are obtained.

[0102] The interaction key feedback data corresponding to the initiative garment interaction data is extracted by the preference feature decision model.

[0103] The interaction key feedback data is input into a decision network layer in the preference feature decision model, and a correlation parameter between the interaction key feedback data and target garment preference feature distribution information in the decision network layer is obtained.

[0104] The garment preference feature distribution corresponding to the target garment preference feature distribution information with the maximum correlation parameter is determined as the initial garment preference feature distribution corresponding to the initiative garment interaction data.

[0105] The garment preference heat map corresponding to the initial garment preference feature distribution is calculated based on the maximum correlation parameter corresponding to the target garment preference feature distribution information.

[0106] In some examples, the determination unit is further configured to:

[0107] extract an interaction data interval of a VR interaction node and an interaction mapping partition of the VR interaction node from the target VR interaction activity;

[0108] obtain a related VR display knowledge point corresponding to the interaction data interval of the VR interaction node, and in the related VR display knowledge point, a first node garment preference feature distribution corresponding to the interaction data interval of the VR interaction node is queried as a first associated garment preference feature distribution, and a first garment preference heat map is configured for the first associated garment preference feature distribution;

[0109] obtain a related VR display knowledge point corresponding to the interaction mapping partition of the VR interaction node, and in the related VR display knowledge point, a second node garment preference feature distribution corresponding to the interaction mapping partition of the VR interaction node is queried as a second associated garment preference feature distribution, and a second garment preference heat map is configured for the second associated garment preference feature distribution.

[0110] determine passive clothing interaction data of the target VR interaction activity based on the first associated clothing preference feature distribution and the second associated clothing preference feature distribution, to obtain a second clothing preference feature distribution corresponding to the passive clothing interaction data;

[0111] determine a clothing preference heat map corresponding to the second clothing preference feature distribution based on the passive clothing interaction data, the first clothing preference heat map and the second clothing preference heat map.

[0112] In summary, the method, system and VR terminal for recommending personalized clothing provided by the present disclosure can obtain a first clothing preference feature distribution corresponding to the active clothing interaction data in the target VR interaction activity by intelligently identifying the active clothing interaction data in the target VR interaction activity, obtain a second clothing preference feature distribution corresponding to the passive clothing interaction data of the target VR interaction activity, determine a clothing preference knowledge graph corresponding to the target VR interaction activity based on the first clothing preference feature distribution and the second clothing preference feature distribution, and obtain a clothing personalized push entity corresponding to the target VR interaction activity through the clothing preference knowledge graph to recommend clothing for the VR interaction activity. In this way, by obtaining a clothing preference knowledge graph for describing the target VR interaction activity and quickly searching for a clothing personalized push entity corresponding to the target VR interaction activity based on the clothing preference knowledge graph, clothing recommendation reliability can be improved.

[0113] The above is only various embodiments of the present disclosure, but the protection scope of the present disclosure is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present disclosure, which should be covered within the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be subject to the protection scope of the claims.

Claims

1. A method of recommending a personalized clothing through a VR terminal, characterized by, The method comprises: obtaining a first clothing preference feature distribution corresponding to active clothing interaction data in a target VR interaction activity; obtaining a second clothing preference feature distribution corresponding to passive clothing interaction data of the target VR interaction activity; determining a clothing preference knowledge graph corresponding to the target VR interaction activity based on the first clothing preference feature distribution and the second clothing preference feature distribution; obtaining a clothing personalized push entity corresponding to the target VR interaction activity based on the clothing preference knowledge graph to recommend clothing for the VR interaction activity, The method comprises: based on the first clothing preference feature distribution and the second clothing preference feature distribution, determining a clothing preference knowledge graph corresponding to the target VR interaction activity, comprising: based on the clothing preference heat map corresponding to the first clothing preference feature distribution and the clothing preference heat map corresponding to the second clothing preference feature distribution, performing attention feature extraction on the first clothing preference feature distribution and the second clothing preference feature distribution to obtain N attention features corresponding to the target VR interaction activity; N is the sum of the number of categories of the first clothing preference feature distribution and the number of categories of the second clothing preference feature distribution; based on the clothing preference heat map corresponding to the N attention features, selecting M attention features from the N attention features as the clothing preference knowledge graph corresponding to the target VR interaction activity, 2. The method of claim 1, wherein, extracting an interaction data interval of a VR interaction node and an interaction mapping partition of the VR interaction node from the target VR interaction activity; obtaining a related VR display knowledge point corresponding to the interaction data interval of the VR interaction node, and configuring a first clothing preference heat map for a first associated clothing preference feature distribution corresponding to the interaction data interval of the VR interaction node in the related VR display knowledge point; obtaining a related VR display knowledge point corresponding to the interaction mapping partition of the VR interaction node, and configuring a second clothing preference heat map for a second associated clothing preference feature distribution corresponding to the interaction mapping partition of the VR interaction node in the related VR display knowledge point; determining passive clothing interaction data of the target VR interaction activity based on the first associated clothing preference feature distribution and the second associated clothing preference feature distribution, and obtaining a second clothing preference feature distribution corresponding to the passive clothing interaction data; determining a clothing preference heat map corresponding to the second clothing preference feature distribution based on the passive clothing interaction data, the first clothing preference heat map, and the second clothing preference heat map. The method further comprises: The target VR interactive activity is acquired, and initiative clothing interaction data in the target VR interactive activity is extracted; preference feature decision of the initiative clothing interaction data is performed through a preference feature decision model, to obtain an initial clothing preference feature distribution corresponding to the initiative clothing interaction data and a clothing preference heat map corresponding to the initial clothing preference feature distribution; An associated clothing preference feature distribution corresponding to the initial clothing preference feature distribution is acquired; The initial clothing preference feature distribution and the associated clothing preference feature distribution are fused, and the fused initial clothing preference feature distribution is determined as a first clothing preference feature distribution corresponding to the initiative clothing interaction data; The clothing preference heat map corresponding to the initial clothing preference feature distribution is determined as a clothing preference heat map corresponding to the first clothing preference feature distribution.

3. The method of claim 2, wherein, The preference feature decision of the initiative clothing interaction data through the preference feature decision model to obtain the initial clothing preference feature distribution corresponding to the initiative clothing interaction data and the clothing preference heat map corresponding to the initial clothing preference feature distribution includes: The interaction key feedback data corresponding to the initiative clothing interaction data is extracted through the preference feature decision model; the interaction key feedback data is input into a decision network layer in the preference feature decision model, to obtain a correlation parameter between the interaction key feedback data and target clothing preference feature distribution information in the decision network layer; the clothing preference feature distribution corresponding to the target clothing preference feature distribution information with the maximum correlation parameter is determined as the initial clothing preference feature distribution corresponding to the initiative clothing interaction data; the clothing preference heat map corresponding to the initial clothing preference feature distribution is calculated based on the maximum correlation parameter corresponding to the target clothing preference feature distribution information.

4. A system for recommending personalized clothing through a VR terminal, characterized by, It includes: The first acquisition unit is configured to acquire a first clothing preference feature distribution corresponding to initiative clothing interaction data in a target VR interactive activity; The second acquisition unit is configured to acquire a second clothing preference feature distribution corresponding to passive clothing interaction data of the target VR interactive activity; The determination unit is configured to determine a clothing preference knowledge graph corresponding to the target VR interactive activity based on the first clothing preference feature distribution and the second clothing preference feature distribution; The clothing recommendation unit is configured to acquire a clothing personalized push entity corresponding to the target VR interactive activity based on the clothing preference knowledge graph, to perform clothing recommendation for the VR interactive activity; The determination unit is specifically configured to: perform attention feature extraction on the first clothing preference feature distribution and the second clothing preference feature distribution based on a clothing preference heat map corresponding to the first clothing preference feature distribution and a clothing preference heat map corresponding to the second clothing preference feature distribution, to obtain N attention features corresponding to the target VR interactive activity; N is the sum of the number of categories of the first clothing preference feature distribution and the number of categories of the second clothing preference feature distribution; based on the clothing preference heat map corresponding to the N attention features, M attention features are selected from the N attention features as the clothing preference knowledge graph corresponding to the target VR interactive activity; The determination unit is further configured to: extract an interaction data interval of a VR interactive node and an interaction mapping partition of the VR interactive node from the target VR interactive activity; Obtain the relevant VR display knowledge point corresponding to the interaction data interval of the VR interactive node, and in the relevant VR display knowledge point, the first node clothing preference feature distribution corresponding to the interaction data interval of the VR interactive node is queried as a first associated clothing preference feature distribution, and the first clothing preference heat map is configured for the first associated clothing preference feature distribution; Obtain the relevant VR display knowledge point corresponding to the interaction mapping partition of the VR interactive node, and in the relevant VR display knowledge point, the second node clothing preference feature distribution corresponding to the interaction mapping partition of the VR interactive node is queried as a second associated clothing preference feature distribution, and the second clothing preference heat map is configured for the second associated clothing preference feature distribution; based on the first associated clothing preference feature distribution and the second associated clothing preference feature distribution, determine passive clothing interaction data of the target VR interactive activity, and obtain a second clothing preference feature distribution corresponding to the passive clothing interaction data; Based on the passive clothing interaction data, the first clothing preference heat map, the second clothing preference heat map, determine the clothing preference heat map corresponding to the second clothing preference feature distribution.

5. The system of claim 4, wherein, The determination unit is further configured to: obtain a target VR interactive activity, and extract active clothing interaction data in the target VR interactive activity; perform preference feature decision on the active clothing interaction data through a preference feature decision model, to obtain an initial clothing preference feature distribution corresponding to the active clothing interaction data and a clothing preference heat map corresponding to the initial clothing preference feature distribution; Obtain the associated clothing preference feature distribution corresponding to the initial clothing preference feature distribution; Fuse the initial clothing preference feature distribution and the associated clothing preference feature distribution, and determine the fused initial clothing preference feature distribution as a first clothing preference feature distribution corresponding to the active clothing interaction data; determine the initial clothing preference feature distribution corresponding to the clothing preference heat map as the clothing preference heat map corresponding to the first clothing preference feature distribution; the preference feature decision model is used to make a preference feature decision on the active clothing interaction data, to obtain the initial clothing preference feature distribution corresponding to the active clothing interaction data and the clothing preference heat map corresponding to the initial clothing preference feature distribution, including: extracting the interaction key feedback data corresponding to the active clothing interaction data by using the preference feature decision model; inputting the interaction key feedback data into a decision network layer in the preference feature decision model, to obtain a correlation parameter between the interaction key feedback data and target clothing preference feature distribution information in the decision network layer; determining the clothing preference feature distribution corresponding to the target clothing preference feature distribution information with the largest correlation parameter as the initial clothing preference feature distribution corresponding to the active clothing interaction data; the clothing preference heat map corresponding to the initial clothing preference feature distribution is calculated based on the largest correlation parameter corresponding to the target clothing preference feature distribution information.

6. A VR terminal, characterized by The method comprises a processor and a machine readable storage medium, the machine readable storage medium stores machine executable instructions, and the processor executes the machine executable instructions to implement the method in any one of claims 1-3.

Images