Virtual model retrieval method and device, equipment, medium and product
Through intelligent matching technology, the search characteristics are determined using CLIP and VIT models, and the input model is directly extracted from the input image, solving the problem of high cost of building a virtual model library and achieving efficient and accurate model search.
Patent Information
- Application Number
- CN202510245342.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-03
- Publication Date
- 2025-07-25
AI Technical Summary
In the prior art, building a virtual model library requires text tags and manual classification, resulting in high cost and low search efficiency.
By responding to user search requests, the input images that search feature matching are determined, and the associated input models are extracted from the model library, and intelligently matched using the CLIP model and VIT image classification model to reduce manual labeling and manual classification steps.
It reduces the cost of model library construction, improves the accuracy and efficiency of virtual model search, and reduces the consumption of human resources.
Smart Images
Figure CN120372032A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the technical field of data processing, and more particularly, to a method, apparatus, device, medium and product for retrieving virtual models. Background Art
[0002] With the rapid development of the game industry, the demand for high-quality art assets in the game industry is increasing day by day. The art assets in the game industry may include virtual models such as virtual characters, virtual vehicles or virtual buildings, and users need to search for the required virtual models in the model library by means of text tags and manual classification, which makes the cost of building the model library relatively high. Summary of the Invention
[0003] An object of an embodiment of the present disclosure is to provide a new technical solution for retrieving virtual models.
[0004] According to a first aspect of the present disclosure, there is provided a method for retrieving a virtual model, the method comprising:
[0005] In response to a retrieval request from a first user, determining a retrieval feature reflected by the retrieval request;
[0006] In a pre-set image library, determining a first input image that matches the retrieval feature;
[0007] In a pre-set model library, extracting a first input model associated with the first input image;
[0008] Sending the first input model to the first user.
[0009] In a possible implementation, the responding to a retrieval request from a first user and determining a retrieval feature reflected by the retrieval request includes:
[0010] In response to a retrieval request from a first user, determining an input type of the retrieval request;
[0011] According to a pre-set processing model, determining a retrieval feature reflected by the retrieval request; wherein the processing model has a corresponding relationship with the input type.
[0012] In a possible implementation, the input type includes a text type, the processing model includes a text processing model, and the text processing model is a CLIP model;
[0013] The determining a retrieval feature reflected by the retrieval request according to a pre-set processing model includes:
[0014] Determining a target semantics that matches the retrieval request through the text processing model and using it as the retrieval feature reflected by the retrieval request.
[0015] In a possible implementation, the input type includes an image type, the processing model includes an image processing model, and the image processing model is a VIT image classification model;
[0016] Determining the retrieval feature reflected by the retrieval request according to the preset processing model includes:
[0017] Determining, through the image processing model, a target image feature that matches the retrieval request and using it as the retrieval feature reflected by the retrieval request.
[0018] In a possible implementation, the method further includes:
[0019] In response to a preview request from the first user for the first input model, changing the first model format of the first input model to a set second model format;
[0020] Performing a set model simplification process on the first input model in the second model format to obtain a processed first input model;
[0021] Displaying the processed first input model to the first user.
[0022] In a possible implementation, the method further includes:
[0023] In response to a trigger task set by a second user, importing a cloud file package;
[0024] When there is an input model to be entered in the file package indicated by the second user that does not belong to the cloud file package, extracting the model feature of the input model to be entered;
[0025] Associatively uploading the input model to be entered and the model feature to the model library.
[0026] According to the second aspect of the present disclosure, there is also provided a retrieval device for a virtual model, including:
[0027] A response module, configured to determine the retrieval feature reflected by the retrieval request in response to a retrieval request from a first user;
[0028] A determination module, configured to determine a first input image that matches the retrieval feature in a preset image library;
[0029] An extraction module, configured to extract a first input model associated with the first input image from a preset model library;
[0030] A sending module, configured to send the first input model to the first user.
[0031] According to a third aspect of the present disclosure, there is also provided a computer system, which includes a processor. When the processor executes program instructions or code, the computer system implements the retrieval method of the virtual model in the first aspect. Exemplarily, the computer system further includes a memory for storing the program instructions or code.
[0032] According to a fourth aspect of the present disclosure, there is also provided a computer-readable storage medium storing a computer program, wherein the computer program is configured to execute the above-mentioned retrieval method of the virtual model when running.
[0033] According to a fifth aspect of the present disclosure, there is also provided a computer program product including a game program. When the game program is executed, the computer is caused to execute the steps of the above-mentioned retrieval method of the virtual model.
[0034] According to a sixth aspect of the present disclosure, there is also provided an electronic device including a memory and a processor. A computer program is stored in the memory, and the processor is configured to execute the above-mentioned retrieval method of the virtual model through the computer program.
[0035] One beneficial effect of the embodiments of the present disclosure is that the retrieval method of the virtual model provided by the embodiments of the present disclosure can respond to the retrieval request of the first user, determine the retrieval features reflected by the retrieval request, and determine the first input image matching the retrieval features, and then extract the first input model associated with the first input image from the model library. In other words, first determine the first input image that meets the retrieval request input by the first user, and then extract the first input model through the first input image. On the one hand, there is no need to set text tags or manual classification, and the first input model can be directly obtained, effectively reducing the construction cost of the model library. On the other hand, the accuracy of searching for the required virtual model can be improved.
[0036] Through the following detailed description of the exemplary embodiments of the present specification with reference to the accompanying drawings, the features and advantages of the embodiments of the present specification will become clear. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] The drawings incorporated in the specification and constituting a part of the specification illustrate the embodiments of the present specification and, together with the description, are used to explain the principles of the embodiments of the present specification.
[0038] Figure 1 FIG. shows a schematic hardware structure diagram of an electronic device that can be used to implement the retrieval method of the virtual model according to the embodiments of the present disclosure;
[0039] Figure 2 FIG. shows a schematic flowchart of the retrieval method of the virtual model according to some embodiments;
[0040] Figure 3 The structural schematic diagram of a retrieval device for a virtual model according to some embodiments is shown;
[0041] Figure 4 The hardware structural schematic diagram of an electronic device according to some embodiments is shown. Detailed implementation manners
[0042] Various exemplary embodiments of the present specification will now be described in detail with reference to the accompanying drawings.
[0043] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way a limitation on the embodiments of the present specification, their applications, or uses.
[0044] It should be noted that like reference numerals and letters denote like items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further discussed in subsequent drawings.
[0045] It should be noted that all actions of obtaining signals, information, or data in the embodiments of the present disclosure are carried out on the premise of complying with the corresponding data protection regulations and policies of the country where the location is located and obtaining the authorization given by the owner of the corresponding device.
[0046] The embodiments of the present disclosure provide a new retrieval scheme for virtual models, which allows a user to provide a retrieval request. By using the retrieval features reflected by the retrieval request, an image that matches the retrieval features is determined, and then a model associated with the image in the model library is extracted. In this way, by first determining a first input image that meets the retrieval request of the first user input, and then extracting a first input model through the first input image. On the one hand, there is no need to set text tags or manually classify, and the required model can be directly obtained, effectively reducing the construction cost of the model library. On the other hand, the accuracy of searching for the required virtual model can be improved.
[0047] Figure 1 The hardware structural schematic diagram of an electronic device that can be used to implement the retrieval method of the virtual model according to the embodiments of the present disclosure is shown.
[0048] The electronic device 1000 is a device capable of running art design software. The art design software can be a local application installed on the electronic device, or a web application, a lightweight application, or a mini-program, etc., which is not limited herein. The electronic device 1000 can be a mobile phone, a tablet computer, a PC, etc., which is not limited herein.
[0049] As Figure 1 shown, the electronic device 1000 may include a processor 1101, a memory 1102, an interface device 1103, a communication device 1104, an output device 1105, an input device 1106, and so on.Figure 1 The hardware configuration shown is for illustrative purposes only and is in no way intended to limit the present disclosure, its applications, or uses.
[0050] The processor 1101 is used to execute a computer program, which can be written in an instruction set of architectures such as x86, Arm, RISC, MIPS, SSE, etc. The memory 1102 includes, for example, ROM (Read-Only Memory), RAM (Random Access Memory), non-volatile memory such as a hard disk, etc. The interface device 1103 includes, for example, a USB interface, a network cable interface, a headphone interface, etc. The communication device 1104 can perform wired or wireless communication, for example. The communication device 1104 may include at least one short-range communication module, for example, any module that performs short-range wireless communication based on short-range wireless communication protocols such as the Hilink protocol, WiFi (IEEE 802.11 protocol), Mesh, Bluetooth, ZigBee, Thread, Z-Wave, NFC, UWB, LiFi, etc. The communication device 1104 may also include a remote communication module, for example, any module that performs WLAN, GPRS, 2G / 3G / 4G / 5G remote communication. The output device 1105 may include, for example, a liquid crystal display screen or a touch display screen, a speaker, etc. The input device 1106 may include, for example, a touch screen, a keyboard, a microphone, various sensors, etc.
[0051] In this embodiment, the memory 1102 of the electronic device 1000 is used to store a computer program, which is used to control the processor 1101 to operate to execute the retrieval method of the virtual model according to any embodiment of the present disclosure.
[0052] Next, taking the electronic device 1000 as an example of the implementation subject, various embodiments of the retrieval method of the virtual model will be described. Figure 1 The various embodiments of the retrieval method of the virtual model will be described by taking the electronic device 1000 as an example of the implementation subject.
[0053] <First Embodiment>
[0054] Figure 2 A retrieval method of a virtual model according to some embodiments is shown. The retrieval method of the virtual model may include the following steps S210 to S240:
[0055] Step S210, in response to a retrieval request from a first user, determine the retrieval features reflected by the retrieval request.
[0056] In this embodiment, the retrieval request may be one, a paragraph, or multiple paragraphs of text, one or multiple pictures, or a video segment, etc.
[0057] In this embodiment, the retrieval features reflected by the retrieval request reflect the key features in text, pictures, or videos. When the retrieval feature is text, the retrieval feature can be a keyword in the text, such as "tree", "car", or "river", etc. When the retrieval feature is a picture or a video, the retrieval feature can be the key features in the picture or video, such as "tree trunk", "wheel", or "pyramid", etc.
[0058] Step S220: In the pre-set image library, determine the first input image that matches the retrieval feature.
[0059] In this embodiment, the image library can be pre-set, and the image library can contain two-dimensional images of different perspectives of the input model. For example, if the retrieval feature is a tree trunk, the first input image with the feature of a tree trunk can be determined in the image library.
[0060] Step S230: In the pre-set model library, extract the first input model associated with the first input image.
[0061] In this embodiment, different input images are all associated with an input model. For example, the model identifier of input model A1 is A1, the image identifier of the top-down two-dimensional image A2 is A2, and the image identifier of the side-view two-dimensional image B1 is B1. Then, both the input model A1 and the top-down two-dimensional image A2 have the identifier "A", so the input model A1 and the top-down two-dimensional image A2 are associated.
[0062] Step S240: Send the first input model to the first user.
[0063] In this embodiment, after the first input model is extracted, the first input model can be sent to the first user so that the first user can use the first input model.
[0064] According to the retrieval method of the virtual model in the first embodiment of the present invention, it solves the problem in the prior art that users need to search for the required virtual model in the model library by means of text labels and manual classification, which makes the cost of building the model library relatively high. Based on this method, in response to the retrieval request of the first user, determine the retrieval features reflected by the retrieval request, determine the first input image that matches the retrieval features, and then extract the first input model associated with the first input image in the model library. In other words, first determine the first input image that meets the retrieval request input by the first user, and then extract the first input model through the first input image. On the one hand, there is no need to set text labels or manual classification, and the first input model can be directly obtained, effectively reducing the cost of building the model library. On the other hand, it can improve the accuracy of searching for the required virtual model.
[0065] <Second Embodiment>
[0066] In this embodiment, in order to be applicable to different retrieval scenarios, the input type of the retrieval request can be determined, and according to different input types, the retrieval features reflected by the retrieval request can be determined, so as to subsequently determine the first input image that matches the retrieval features.
[0067] In these embodiments, compared with the above first embodiment, step S220 may include the following step S310 and step S320:
[0068] Step S310, in response to the retrieval request of the first user, determine the input type of the retrieval request.
[0069] In this embodiment, when the first user issues a retrieval request, the input format of the retrieval request can be determined, and through this input format, the input type of the retrieval request can be determined. For example, if the input format is a picture format, then the input type of the retrieval request is an image type.
[0070] Step S320, according to the preset processing model, determine the retrieval features reflected by the retrieval request; wherein, the processing model has a corresponding relationship with the input type.
[0071] In this embodiment, by setting different processing models, different retrieval requests can be applied to determine the retrieval features reflected by the retrieval request, so as to subsequently determine the first input image that matches the retrieval features.
[0072] <Third Embodiment>
[0073] In this embodiment, in order to implement a retrieval scenario applicable to text input, the input type may include a text type, the processing model includes a text processing model, and the text processing model is a CLIP model. Furthermore, through this text processing model, the retrieval features in the text can be determined, so as to implement the determination of the first input image that matches the retrieval features.
[0074] In these embodiments, compared with the above second embodiment, step S320 may include the following step S410:
[0075] Step S410, through the text processing model, determine the target semantics that match the retrieval request and use it as the retrieval features reflected by the retrieval request.
[0076] In this embodiment, the CLIP model (Contrastive Language-Image Pre-training) realizes cross-modal information retrieval by jointly training text and images. The CLIP model is adapted to Chinese natural language, and for Chinese natural language, the CLIP model is localized and optimized to better support the Chinese context. Specifically, it includes retraining on the Chinese dataset, updating the word embedding model and the language model to enable it to more accurately understand Chinese semantics. The CLIP model provides more intelligent semantic matching capabilities, effectively improving the accuracy and efficiency of retrieval. In addition, the adaptive learning ability of the CLIP model enables the system to enhance its understanding of user intentions during continuous use.
[0077] In this embodiment, the CLIP model can be a model optimized for Chinese. The CLIP model can analyze the features of virtual models, texture maps, and 3D scene screenshots, and extract the semantic features of the natural language input by the user, that is, the target semantics matching the retrieval request. Through the CLIP, it is possible to achieve intelligent understanding of Chinese search keywords and accurately match relevant virtual models, effectively avoiding the inability to search through natural language and the poor search effect in the Chinese environment. And it effectively reduces the situation where a large amount of manpower needs to be spent in advance to label virtual models when they are put into the library, and at the same time solves the problems of the retrieval tool consuming a large amount of manpower and the error-prone manual labeling.
[0078] <Fourth Embodiment>
[0079] In this embodiment, in order to implement a retrieval scenario applicable to image input, the input type includes the image type, and the processing model includes an image processing model. The image processing model is a VIT image classification model. Furthermore, through this image processing model, the retrieval features in the image can be determined, and then the first input image matching the retrieval features can be determined.
[0080] In these embodiments, compared with the second embodiment above, the step S320 may include the following step S510:
[0081] Step S510, determine the target image features matching the retrieval request through the image processing model and use them as the retrieval features reflected by the retrieval request.
[0082] In this embodiment, the VIT image classification model (Vision Transformer) is a deep learning model that applies the Transformer architecture to computer vision tasks.
[0083] In this embodiment, by means of the ViT image classification model, the target image features in the retrieval request can be accurately extracted and quickly matched with the input images in the subsequent image library. Moreover, the VIT image classification model can support various picture formats and can process input pictures from complex scenes and various lighting conditions to achieve high-precision search capabilities, thereby enabling the first user to retrieve similar virtual models through retrieval requests in the form of uploaded pictures or videos.
[0084] In this embodiment, the retrieval request may include game screen screenshots and concept sketches. For this retrieval request, the target image features matching the retrieval request can be determined to meet the requirement of searching for similar virtual models.
[0085] <Fifth Embodiment>
[0086] In this embodiment, to facilitate the user to view whether the obtained first input model is the model they need, the model format of the first input model can be changed so that the first input model can be previewed.
[0087] In these embodiments, compared with the above first embodiment, after step S230, the method further includes the following steps S610 to S630:
[0088] Step S610, in response to the preview request of the first user for the first input model, change the first model format of the first input model to the set second model format.
[0089] In this embodiment, the set second model format can be a specified model format or a model format supported by the user locally.
[0090] Step S620, perform set model simplification processing on the first input model in the second model format to obtain the processed first input model.
[0091] In this embodiment, the model simplification processing can be to simplify the content of the first input model. For example, delete the filling of the first input model, delete the texture of the first input model, etc.
[0092] Step S630, display the processed first input model to the first user.
[0093] In this embodiment, through the preview of the first input model, it is convenient for the user to view whether the obtained first input model is the model they need.
[0094] <Sixth Embodiment>
[0095] In this embodiment, in order to improve the efficiency of virtual model warehousing, a cloud file package can be directly imported, the cloud file package is compared with the file package indicated by the second user, and the model to be entered whose indicated file package is different from the cloud file package is uploaded to the model library, so as to automatically complete the upload of the virtual model and effectively improve the management efficiency of art assets such as virtual models.
[0096] In these embodiments, compared with the first embodiment above, step S2501 may include the following steps S710 to S730:
[0097] Step S710, in response to the trigger task set by the second user, import the cloud file package.
[0098] In this embodiment, the trigger task may include an active trigger task and a passive trigger task. The active trigger task is, for example, an instruction for the second user to actively issue an import of the cloud file package, and the passive trigger task is, for example, an instruction for the cloud file package to be passively issued regularly.
[0099] Step S720, when there is a model to be entered in the file package indicated by the second user that does not belong to the cloud file package, extract the model features of the model to be entered.
[0100] In this embodiment, the cloud file package may include each entered model, or may include two-dimensional images of each perspective of each entered model.
[0101] In this embodiment, before step S720, step S719 may be included. Step S719 may be: determining the model to be entered in the file package indicated by the second user, and determining the two-dimensional image of the specified perspective of the model to be entered; comparing the similarity between the two-dimensional image of the specified perspective and the two-dimensional image in the cloud file package, and when the similarity between the two-dimensional image of the specified perspective and the two-dimensional image in the cloud file package does not exceed the set threshold, determining that there is a model to be entered in the file package indicated by the second user that does not belong to the cloud file package.
[0102] Step S730, associatively upload the model to be entered and the model features to the model library.
[0103] In this embodiment, the model features may be two-dimensional images of different perspectives of the model to be entered. After step S720, step S721 may be included. Step S721 may be: uploading the model features to the image library. In other words, by associating the model to be entered and the model features, it is possible to query the associated entered model through the entered image.
[0104] <Device Embodiment>
[0105] Figure 3The composition structure diagram of a retrieval device for a virtual model according to an embodiment of the present disclosure is shown. As Figure 3 shown, the retrieval device 300 for the virtual model includes a response module 310, a determination module 320, an extraction module 330, and a sending module 340.
[0106] The response module 310 is configured to determine the retrieval features reflected by the retrieval request in response to the retrieval request of the first user;
[0107] The determination module 320 is configured to determine a first input image that matches the retrieval features in a preset image library;
[0108] The extraction module 330 is configured to extract a first input model associated with the first input image from a preset model library;
[0109] The sending module 340 is configured to send the first input model to the first user.
[0110] In some embodiments, the response module 310 is further configured to determine the input type of the retrieval request in response to the retrieval request of the first user; determine the retrieval features reflected by the retrieval request according to a preset processing model; wherein, the processing model has a corresponding relationship with the input type.
[0111] In some embodiments, the response module 310 is further configured to determine a target semantics that matches the retrieval request through a text processing model and use it as the retrieval features reflected by the retrieval request.
[0112] In some embodiments, the response module 310 is further configured to determine target image features that match the retrieval request through an image processing model and use it as the retrieval features reflected by the retrieval request.
[0113] In some embodiments, the retrieval device 300 for the virtual model further includes a display module, and the display module is configured to change the first model format of the first input model to a set second model format in response to a preview request of the first user for the first input model; perform a set model simplification process on the first input model in the second model format to obtain a processed first input model; display the processed first input model to the first user.
[0114] In some embodiments, the retrieval device 300 for the virtual model further includes an upload module, and the upload module is configured to import a cloud file package in response to a trigger task set by a second user; extract the model features of a to-be-input model when the to-be-input model that does not belong to the cloud file package exists in the file package indicated by the second user; associate and upload the to-be-input model and the model features to the model library.
[0115] <Device Embodiment>
[0116] Figure 4 The schematic diagram of the hardware structure of an electronic device according to some other embodiments is shown. As Figure 4 shown, the electronic device 400 includes a processor 410 and a memory 420. The memory 420 is used to store a computer program, and the computer program is used to control the operation of the processor 410 to control the electronic device 400 to execute the retrieval method of the virtual model according to any embodiment of the present disclosure.
[0117] An embodiment of the present disclosure also provides a computer-readable storage medium, which stores a computer program, and the computer program realizes the retrieval method of the virtual model according to any embodiment of the present disclosure when executed by a processor.
[0118] An embodiment of the present disclosure also provides a computer program product, which includes a computer program or instruction, and the computer program or instruction realizes the retrieval method of the virtual model according to any embodiment of the present disclosure when executed by a processor.
[0119] Each embodiment in this specification is described in a progressive manner. The same or similar parts among the embodiments can be referred to each other, and the differences between each embodiment and other embodiments are emphasized. In particular, for the device and equipment embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiments.
[0120] The above describes specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be executed in a different order than in the embodiments and still achieve the desired result. Additionally, the processes depicted in the figures do not necessarily require the particular order or sequential order shown to achieve the desired result. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0121] Embodiments of this specification can be devices, methods, and / or computer program products. The computer program product can include a computer-readable storage medium having thereon computer-readable program instructions for causing a processor to implement various aspects of the embodiments of this specification.
[0122] A computer-readable storage medium can be a tangible device that can hold and store instructions for use by an instruction execution device. A computer-readable storage medium may be, for example—but not limited to—an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disc (DVD), a memory stick, a floppy disk, a mechanically encoded device, such as a punched card or raised structures in grooves storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage medium used herein is not construed as an instantaneous signal itself, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., an optical pulse through an optical fiber cable), or an electrical signal transmitted through a wire.
[0123] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to respective computing / processing devices, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network may include a copper transmission cable, an optical fiber transmission, a wireless transmission, a router, a firewall, a switch, a gateway computer, and / or an edge server. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in each computing / processing device.
[0124] The computer program instructions for performing the operations of the embodiments of this specification may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages such as Smalltalk, C++, etc., and conventional procedural programming languages such as the "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the first user computer, partially on the first user computer, executed as a stand-alone software package, partially on the first user computer and partially on a remote computer, or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the first user computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, by using the state information of the computer-readable program instructions to customize an electronic circuit, such as a programmable logic circuit, a field-programmable gate array (FPGA), or a programmable logic array (PLA), the electronic circuit can execute the computer-readable program instructions to implement various aspects of the embodiments of this specification.
[0125] Aspects of the embodiments of this specification are described herein with reference to the flowcharts and / or block diagrams of methods, apparatus (devices), and computer program products according to the embodiments of this specification. It should be understood that each block of the flowcharts and / or block diagrams, and the combinations of blocks in the flowcharts and / or block diagrams, can be implemented by computer-readable program instructions.
[0126] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that when these instructions are executed by the processor of the computer or other programmable data processing apparatus, a device is created that implements the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions cause the computer, programmable data processing apparatus, and / or other devices to operate in a specific manner. Thus, the computer-readable medium storing the instructions includes a manufacture that includes instructions for implementing various aspects of the functions / actions specified in one or more blocks of the flowchart and / or block diagram.
[0127] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device, causing a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process such that the instructions executed on the computer, other programmable data processing apparatus, or other device implement the functions / acts specified in one or more boxes of the flowchart and / or block diagram.
[0128] The flowcharts 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 specification. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a portion of an instruction, which contains one or more executable instructions for implementing the specified logical function. In some alternative implementations, the functions noted in the blocks may occur out of the order noted in the figures. For example, two consecutive blocks may in fact be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending on the functionality involved. It should also be noted that each block of the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented by a dedicated hardware-based system that performs the specified functions or acts, or by a combination of dedicated hardware and computer instructions. As is well known to those skilled in the art, implementation by hardware, implementation by software, and implementation by a combination of software and hardware are equivalent.
[0129] The embodiments of the present specification have been described above. The above description is exemplary and not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art in the technical field without departing from the scope of the described embodiments. The choice of terms used herein is intended to best explain the principles of the embodiments, the practical application, or the improvement of the technology in the market, or to enable other ordinary skilled persons in the technical field to understand the embodiments disclosed herein.
Claims
1. A retrieval method for virtual models, the method comprising: Responding to a retrieval request from a first user, determining the retrieval features reflected by the retrieval request; In a preset image library, determining a first input image that matches the retrieval features; In a preset model library, extracting a first input model associated with the first input image; Sending the first input model to the first user.
2. The method according to claim 1, wherein The step of responding to a retrieval request from a first user and determining the retrieval features reflected by the retrieval request includes: Responding to a retrieval request from a first user, determining the input type of the retrieval request; According to a preset processing model, determining the retrieval features reflected by the retrieval request; wherein, the processing model has a corresponding relationship with the input type.
3. The method according to claim 2, wherein The input type includes a text type, the processing model includes a text processing model, and the text processing model is a CLIP model; The step of determining the retrieval features reflected by the retrieval request according to a preset processing model includes: Through the text processing model, determining a target semantics that matches the retrieval request and using it as the retrieval features reflected by the retrieval request.
4. The method according to claim 2, wherein The input type includes an image type, the processing model includes an image processing model, and the image processing model is a VIT image classification model; The step of determining the retrieval features reflected by the retrieval request according to a preset processing model includes: Through the image processing model, determining a target image feature that matches the retrieval request and using it as the retrieval features reflected by the retrieval request.
5. The method according to claim 1, wherein, The method further includes: Responding to a preview request from the first user for the first input model, changing the first model format of the first input model to a set second model format; Performing a set model simplification process on the first input model in the second model format to obtain a processed first input model; Showing the processed first input model to the first user.
6. The method according to claim 1, wherein, The method further includes: Responding to a trigger task set by a second user, importing a cloud file package; When there is a model to be input in the file package indicated by the second user that does not belong to the cloud file package, extracting the model features of the model to be input; Associatively uploading the model to be input and the model features to the model library.
7. A retrieval device for a virtual model, wherein, including: A response module, configured to respond to a retrieval request from a first user and determine the retrieval features reflected by the retrieval request; A determination module, configured to determine a first input image that matches the retrieval features in a preset image library; An extraction module, configured to extract a first input model associated with the first input image in a preset model library; A sending module, configured to send the first input model to the first user.
8. An electronic device, wherein, Including a memory and a processor, the memory is used to store a computer program, and the processor is used to execute the method steps according to any one of claims 1 to 6 under the control of the computer program.
9. A computer-readable storage medium, the computer-readable storage medium comprising a stored computer program, wherein, When the computer program is run, it executes the method steps according to any one of claims 1 to 6.
10. A computer program product, characterized in that, Comprising a computer program which, when executed by a processor, implements the method according to any one of claims 1 to 6.