Spatial layout information generation method and device, storage medium and program product

Abstract

The embodiment of the invention provides a spatial layout information generation method and device, a storage medium and a program product. In the embodiment of the application, the spatial layout information is generated for the target spatial object by means of the understanding ability of the spatial layout generation model for the article layout knowledge. The spatial layout generation model corresponds to a target spatial layout generation process, and at least part of execution steps in the process depend on layout reference information retrieved by a layout knowledge base; and calling a spatial layout generation model, and automatically executing a target spatial layout generation process according to the layout reference information to generate spatial layout information. Through combination of the layout reference information and the target spatial layout generation process, the reasoning capability of the spatial layout generation model in layout generation can be remarkably improved, the generation efficiency and the generation quality of the spatial layout information are improved, and the generation cost is reduced. Furthermore, layout optimization processing is carried out for the spatial layout information, and the reasonability and regularity of the spatial layout information are further improved.

Description

Technical Field

[0001] The present application relates to the field of smart home improvement technology, and in particular to a method, device, storage medium and program product for generating spatial layout information. Background Art

[0002] In the field of e-commerce, furniture merchants hope to display their furniture products in real indoor home decoration scenes. Merchants have the need to display multiple furniture products in an indoor space, which requires designers to design a variety of home matching scenes. For example, determine which furniture to place in the space and plan the specific location, orientation and other spatial information of each piece of furniture. In the traditional interior home decoration design industry, relying solely on manual design to generate a large number of furniture matching scenes has the problems of high cost and low efficiency. Summary of the invention

[0003] Multiple aspects of the present application provide a method, device, storage medium and program product for generating spatial layout information, so as to improve the quality and efficiency of generating spatial layout information and reduce the cost.

[0004] An embodiment of the present application provides a method for generating spatial layout information, including: obtaining description information of a target spatial object for which spatial layout information is to be generated; obtaining a target spatial layout generation process corresponding to a spatial layout generation model, wherein at least some execution steps in the target spatial layout generation process depend on layout reference information, and different execution steps depend on different layout reference information; according to the description information of the target spatial object, retrieving from a layout knowledge base the layout reference information on which at least some execution steps depend, the layout reference information including item attribute information or layout rule knowledge information as a constraint condition for the execution steps; calling the spatial layout generation model, and executing the execution steps in the target spatial layout generation process according to the layout reference information to obtain the spatial layout information of the target spatial object; wherein the spatial layout information is used to describe target items suitable for being placed in the target spatial object and the placement position relationships between the target items.

[0005] An embodiment of the present application also provides an electronic device, comprising: a memory and a processor; the memory is used to store a computer program; the processor is coupled to the memory and is used to execute the computer program to implement each step in the spatial layout generation method provided in the embodiment of the present application.

[0006] The embodiment of the present application also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor implements each step in the spatial layout generation method provided in the embodiment of the present application.

[0007] The embodiment of the present application also provides a computer program product, including a computer program / instruction. When the computer program / instruction is executed by a processor, the processor is caused to implement each step in the spatial layout generation method provided in the embodiment of the present application.

[0008] In an embodiment of the present application, with the help of the spatial layout generation model's ability to understand the knowledge of object layout, the layout reference information retrieved from the layout knowledge base and the spatial layout generation process are combined, which can significantly improve the reasoning ability of the spatial layout generation model in layout generation. The spatial layout generation process is automatically executed through the spatial layout generation model to generate spatial layout information corresponding to the target spatial object, thereby improving the generation efficiency and quality of the spatial layout information and reducing the generation cost.

[0009] Further optionally, in an embodiment of the present application, at least one layout principle may be adopted to perform layout optimization processing on the spatial layout information to fine-tune the layout of the spatial layout information, thereby improving the spatial rationality and layout regularity of the spatial layout information. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0011] Figure 1 A schematic diagram of a flow chart of a method for generating spatial layout information provided by an exemplary embodiment of the present application;

[0012] Figure 2a A schematic diagram of a target structured prompt word and a layout knowledge base provided for an exemplary embodiment of the present application;

[0013] Figure 2b A schematic diagram of a spatial layout information generation process provided by an exemplary embodiment of the present application;

[0014] Figure 3a and Figure 3b A schematic diagram of the plane layout information and three-dimensional layout information corresponding to a "bedroom" provided in an exemplary embodiment of the present application;

[0015] Figure 4a and Figure 4b A schematic diagram of the plane layout information and three-dimensional layout information corresponding to another "bedroom" provided in an exemplary embodiment of the present application;

[0016] Figure 5a and Figure 5b A schematic diagram of the plane layout information and three-dimensional layout information corresponding to a "living room" provided by an exemplary embodiment of the present application;

[0017] Figure 6 A schematic diagram of the structure of a spatial layout information generating device provided by an exemplary embodiment of the present application;

[0018] Figure 7 A schematic structural diagram of an electronic device provided for an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0020] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0021] The various models involved in this application (including but not limited to language models or large models) are in compliance with relevant laws and standards.

[0022] At present, whether it is a furniture merchant selling furniture, a personal home furniture arrangement, or a designer providing home service, there is a need to generate spatial layout information for spatial objects. However, in the traditional interior design industry, relying solely on manual design to generate a large number of furniture matching scenes has the problems of high cost and low efficiency.

[0023] In response to the problems faced by the above-mentioned scenarios, the embodiment of the present application first proposes a method A1 for generating spatial layout information based on a large language model. The large language model can be a large language model based on in-context learning technology (also often referred to as "few-shot learning"), for example, a pre-trained large language model. In this solution, a sample set is input into a large language model, and each sample includes a sample prompt word and sample spatial layout information. The large language model understands the task it needs to complete through the sample set. During the reasoning process, the large language model can perform a reasoning process for the prompt word input by the user to obtain spatial layout information. However, the effect of generating spatial layout information in the test environment and the actual environment is not very ideal. For example, the generated spatial layout information is consistent with the original layout in the sample set, and it is difficult to break through the sample set, and it is impossible to produce spatial layout information that is closer to the user's intention.

[0024] In view of the above-mentioned deficiencies, the inventors of the present application continued to analyze and study, and proposed another spatial layout information generation method A2. With the help of the spatial layout generation model's ability to understand the knowledge of object layout, the layout reference information retrieved from the layout knowledge base and the spatial layout generation process are combined, which can significantly improve the reasoning ability of the spatial layout generation model in layout generation. The spatial layout generation process is automatically executed through the spatial layout generation model to generate spatial layout information corresponding to the target spatial object, thereby improving the generation efficiency and quality of the spatial layout information and reducing the generation cost.

[0025] Further optionally, the layout reference information includes object layout reference description information, which constrains the spatial layout of the objects in the form of description, mainly describing the positional relationship between the objects and the walls, doors and other structures of the target space object, as well as the positional relationship between the objects, rather than strictly limiting the absolute position information of the objects, thereby making the spatial layout generation model more flexible in generating spatial layout information.

[0026] Further optionally, personalized layout description information provided by the user can be introduced into the spatial layout generation process, so that the user can guide the generation of the spatial layout according to his or her own needs, so that the spatial layout generation model can output spatial layout information that is closer to the user's intention.

[0027] Further optionally, in an embodiment of the present application, at least one layout principle may be adopted to perform layout optimization processing on the spatial layout information to fine-tune the layout of the spatial layout information, thereby improving the spatial rationality and layout regularity of the spatial layout information.

[0028] The following is a detailed description of method A2 provided in an embodiment of the present application in conjunction with the accompanying drawings.

[0029] Figure 1 The following is a flow chart of a method for generating spatial layout information provided by an exemplary embodiment of the present application. Figure 1 As shown, the method includes:

[0030] 101. Obtain description information of a target space object for which spatial layout information is to be generated;

[0031] 102. Obtain a target space layout generation process corresponding to the space layout generation model, wherein at least some execution steps in the target space layout generation process depend on layout reference information, and different execution steps depend on different layout reference information;

[0032] 103. Retrieve, from a layout knowledge base, at least some of the layout reference information that the execution steps depend on, based on the description information of the target space object, wherein the layout reference information includes item attribute information or layout rule knowledge information that is a constraint condition for the execution steps;

[0033] 104. Call the spatial layout generation model, and execute the execution steps in the target spatial layout generation process according to the layout reference information to obtain the spatial layout information of the target spatial object; wherein the spatial layout information is used to describe the target objects suitable for being placed in the target spatial object and the placement position relationship between the target objects.

[0034] In this embodiment, the application scenario of the spatial layout information generation method is not limited. For example, if a merchant wants to place the furniture for sale in a suitable house environment to enhance the sense of entity, the merchant can input the description information of the house and generate the spatial layout information of the house according to the spatial layout information generation method. The spatial layout information includes the size information of each piece of furniture, and the corresponding furniture is selected from the furniture according to the size information; three-dimensional rendering is performed according to the selected furniture and the spatial layout information to obtain three-dimensional spatial layout information to display the furniture in a three-dimensional scene. For another example, if a user wants to buy furniture on an e-commerce APP, the user can upload the description information of the house, and the e-commerce APP generates the spatial layout information of the house according to the spatial layout information generation method, and selects suitable furniture from the furniture library according to the spatial layout information; a three-dimensional scene is generated according to the selected furniture, and the three-dimensional scene includes furniture products. The user can view the three-dimensional scene and purchase the furniture products in the three-dimensional scene.

[0035] In this embodiment, the target space object is not limited. For example, in a home environment, the target space object may include but is not limited to: living room, kitchen, bathroom, bedroom and balcony, etc. For another example, in a shopping mall environment, the target space object may include but is not limited to: clothing store, shoe store, restaurant and snack shop, etc. For another example, in an office scene, the target space object may include but is not limited to: conference room, finance room and office.

[0036] In this embodiment, the target space object refers to the space object for which the spatial layout information is to be generated. The spatial layout information is used to describe the target items that are suitable for being placed in the target space object and the placement position relationships between the target items. That is, the spatial layout information generation method of the embodiment of the present application can determine the target items that are suitable for being placed in the target space object, as well as determine the placement position relationships between the target items.

[0037] Among them, according to the different target space objects, the target items suitable for placement in the target space objects are also different. For example, if the target space object is implemented as a living room, the target items may include but are not limited to: sofas, coffee tables, televisions, chandeliers, cabinet air conditioners. For another example, if the target space object is implemented as a bathroom, the target items may include but are not limited to: toilets, sinks, water heaters, and washing machines. For another example, if the target space object is implemented as an office, the target items may include but are not limited to: desks, office chairs, lockers, water dispensers, and potted plants. For another example, if the target space object is implemented as a clothing store in a shopping mall, the target items may include but are not limited to: shelves, fitting mirrors, fitting rooms, counters, and sofas.

[0038] In this embodiment, the description information of the target space object for which the spatial layout information is to be generated may be obtained. The description information of the target space object may include, but is not limited to, the spatial attribute information and layout type information of the target space object. The spatial attribute information of the target space object may include, but is not limited to, the spatial type information and spatial dimension information of the target space object. For example, the spatial dimension information may include, for example, the length and width of the target space object, and further, the shape of the target space object (such as a cube or a cuboid, etc.), and the door and window position information of the target space object. For example, the spatial type information of the target space object may include, for example, a living room, a kitchen, an office, and a conference room.

[0039] In this embodiment, spatial layout information can be generated for the target spatial object through a spatial layout generation model. Any model that can generate spatial layout information for a target spatial object is applicable to the embodiments of the present application. For example, the spatial layout generation model can adopt but is not limited to: a large language model (LLM), such as a general language generation model, a language dialogue model, or a deformation model of the above models.

[0040] In this embodiment, the spatial layout generation model corresponds to a target spatial layout generation process, and the target spatial layout generation process includes: multiple execution steps with a logical order. For example, the target spatial layout generation process can be generated through a chain of thought (CoT). Specifically, CoT can introduce prior knowledge (such as background information) and common sense reasoning, and gradually expand the task of generating spatial layout information, converting the task of generating spatial layout information into multiple execution steps with a logical order, and retaining the background information of the previous execution step in each execution step, which can assist the spatial layout generation model to generate spatial layout information more reasonably and accurately.

[0041] In this embodiment, at least some of the execution steps in the target space layout generation process rely on layout reference information. For example, all the execution steps included in the target space layout generation process can rely on layout reference information, or a part of multiple execution steps rely on layout reference information. Wherein, different execution steps rely on different layout reference information. Wherein, the layout reference information includes knowledge information related to the layout of items required by the spatial layout generation model to generate spatial layout information. The layout reference information may include item attribute information or layout rule knowledge information as a constraint condition for the execution step. For example, the layout rule knowledge information includes prior knowledge information and empirical knowledge information. Wherein, the layout reference information may be description information, such as text description information or voice description information. Preferably, the layout reference information is text description information. Wherein, usually, a three-dimensional sample room or a two-dimensional layout design drawing is used as reference information for the spatial layout. However, the position information of each item in the spatial object is strictly limited in the three-dimensional sample room or the two-dimensional layout design drawing, resulting in poor flexibility in generating spatial layout information by the spatial layout generation model, and the generated spatial layout information is similar or identical to the existing spatial layout, and the innovation is low. In the embodiment of the present application, the layout reference information is implemented as descriptive information, and the position information of each object in the spatial object is not strictly limited. The layout reference information can describe the relative position relationship between objects and the relative position relationship between objects and the target spatial object, thereby providing a reference for the spatial layout generation model to generate spatial layout information, so that the spatial layout generation model can adaptively generate spatial layout information that meets user needs based on different spatial objects, thereby improving the flexibility and creativity of the spatial layout generation model in generating spatial layout information.

[0042] Among them, the item attribute information may include but is not limited to: item type information, item size information, and item orientation information. Depending on the different spatial type information of the target spatial object, the item type information is also different. For example, if the target spatial object is a kitchen, the item type information may include but is not limited to: gas stoves, range hoods, vegetable sinks, dishwashers, rice cookers, ovens, wall cabinets, and top cabinets. For another example, if the target spatial object is a bedroom, the item type information may include but is not limited to: beds, air conditioners, dressing tables, wardrobes, bedside tables, and hanging paintings. Among them, the item size information includes the size reference range of each item, and the item orientation information may be toward other items, or toward structures such as walls, windows, or doors in the spatial object.

[0043] The layout rule knowledge information may include, but is not limited to, item layout description information and space calculation rule knowledge information. Space calculation rule knowledge information is a kind of calculation rule information used to guide how to reasonably and effectively arrange items in a space object. Space calculation rule information may include, but is not limited to, layout rules (such as safety distance and passage space), space utilization calculation (characterizing comfort), visual balance principles, and ergonomics.

[0044] Among them, the item layout description information is a preset description information for the items in the spatial object and the placement relationship between the items, which can also be called an item layout description template. The item layout description information can be the description information set in advance by the designer. It should be noted that the item layout description information here is a description information, which is different from the three-dimensional sample room or two-dimensional layout design drawing pre-designed by the designer. The item layout description information constrains the spatial layout of the items in the form of description, mainly describing the positional relationship between the items and the walls, doors and other structures of the target spatial object, as well as the positional relationship between the items, rather than strictly limiting the absolute position of the items, so that the spatial layout generation model is more flexible in generating spatial layout information.

[0045] The following is an exemplary description of the item layout corresponding to a "living room in a straight line layout", but is not limited thereto.

[0046] a. The main sofa is best placed in the center of the main wall (assuming the main wall is on the south side) and close to the main wall, with the main sofa facing the center of the room.

[0047] b. The TV cabinet or main cabinet should be placed opposite the main sofa and against the wall opposite the main sofa, with the front direction of the TV cabinet facing the center of the room.

[0048] c. Place a coffee table in front of the main sofa, with a certain distance (45cm-70cm) between the coffee table and the main sofa, and the direction of the coffee table is towards the center of the room.

[0049] d. Place a carpet under the coffee table, with the center of the carpet coinciding with the center of the coffee table, and the carpet is very large.

[0050] In this embodiment, considering that the amount of knowledge in the layout knowledge base is large, if the entire layout knowledge base is used as layout reference information, the efficiency of generating spatial layout information will be reduced. Therefore, the retrieval-augmented generation (RAG) technology is introduced, that is, partial knowledge information for the target spatial object is selected from the layout knowledge base as layout reference information. On the one hand, the amount of calculation for generating spatial layout information is reduced, and on the other hand, more targeted knowledge is selected as layout reference information to assist in generating spatial layout information and improve the accuracy of generating spatial layout information. Specifically, according to the description information of the target spatial object, at least part of the layout reference information that each execution step depends on is retrieved from the layout knowledge base. Among them, when the layout reference information that different execution steps depend on is different, the corresponding layout reference information can be retrieved from different layout knowledge bases.

[0051] In this embodiment, a spatial layout generation model is called, and in the spatial layout generation model, the execution steps in the target spatial layout generation process are executed according to the layout reference information to obtain the spatial layout information of the target spatial object. Among them, the spatial layout information includes target items suitable for being placed in the target spatial object and the placement position relationship between the target items. Specifically, the spatial layout information includes the attribute description information of the target item, the relative position relationship between the target item and the target spatial object, and the placement position relationship between the target items. The number of target items is one or more. Among them, the attribute description information of the target item includes: target type information, target size information, target position information, and target orientation information of the target item. The target spatial object includes: structures such as walls, windows, doors, and ceilings, and the position relationship between the target item and the target spatial object includes: the position relationship between the target item and each structure included in the target spatial object. For example, when the target item is a sofa, the sofa fits the wall C1 of the target spatial object; when the target object is a TV, the TV is hung on the wall C2 of the target spatial object; when the target object is a chandelier, the chandelier is located on the ceiling C3 of the target spatial object. The placement relationship between the target objects may include but is not limited to: the first object is located above, below, to the left, to the right, in front of or behind the second object, etc. For example, in a living room, the coffee table is located above the carpet, the sofa is located behind the coffee table, and the painting is located above the sofa, etc.

[0052] In an embodiment of the present application, spatial layout information is generated for a target space object with the help of the spatial layout generation model's ability to understand the knowledge of object layout. The spatial layout generation model corresponds to a target space layout generation process, in which at least some of the execution steps rely on the layout reference information retrieved from the layout knowledge base; the spatial layout generation model is called to automatically execute the target space layout generation process according to the layout reference information to generate the spatial layout information. Among them, the combination of layout reference information and target space layout generation process can significantly improve the reasoning ability of the spatial layout generation model in layout generation, and obtain reasonable, rich and user-friendly spatial layout information. Furthermore, layout optimization processing is performed on the spatial layout information, and layout fine-tuning is performed on the spatial layout information to further improve the spatial rationality and layout regularity of the spatial layout information.

[0053] In an optional embodiment, a structured prompt word template can be designed in advance for the spatial layout generation model, and the structured prompt word template includes: an initial spatial layout generation process, the initial spatial layout generation process includes multiple execution steps, and the multiple execution steps have a logical execution order, wherein some execution steps have input-output dependencies. For example, the reasoning results of some execution steps can be used as input for subsequent execution steps. Wherein, at least some of the execution steps include: retrieval guide information. Retrieval guide information is a kind of guidance information that assists in retrieving layout reference information from a layout knowledge base. An implementation method of retrieving layout reference information that at least some of the execution steps depend on from a layout knowledge base based on the description information of the target space object, including: under the guidance of the retrieval guide information, according to the description information of the target space object, retrieving the layout reference information that at least some of the execution steps depend on from the layout knowledge base.

[0054] Among them, under the guidance of the retrieval guidance information, the layout reference information related to the target space object can be quickly retrieved, the layout experience can be quickly obtained, the layout design level can be improved, and the invalid search time can be reduced.

[0055] Optionally, the layout knowledge base can be classified. For example, the layout knowledge base can include multiple types, such as an item attribute knowledge base and a layout rule knowledge base. Further, the item attribute knowledge base can include but is not limited to: an item type knowledge base, an item size knowledge base, and an item orientation knowledge base, and the layout rule knowledge base can include but is not limited to: a layout description knowledge base and a spatial calculation knowledge base. For a detailed introduction to each layout knowledge base, please refer to the subsequent embodiments, which will not be repeated here. Based on this, an implementation method of retrieving the layout reference information on which at least some execution steps depend from the layout knowledge base according to the description information of the target space object under the guidance of the retrieval guidance information includes: selecting a target layout knowledge base adapted to the target execution step information from multiple layout knowledge bases according to the target execution step information to which the retrieval guidance information belongs, and the target execution step belongs to at least some execution steps; selecting target description information adapted to the content information from the description information according to the content information of the retrieval guidance information; retrieving the item attribute information or layout rule knowledge information adapted to the target description information in the target layout knowledge base as the layout reference information on which the target execution step depends. Among them, by retrieving the execution steps to which the guidance information belongs and the content information of the guidance information, an automated retrieval process can be realized, and a suitable layout knowledge base and description information can be quickly found for the target execution step, thereby saving time for manual search and improving retrieval efficiency. In addition, by accurately selecting the target description information based on the content information and matching it with the target layout knowledge base, the obtained layout reference information can be made to more accurately meet the actual needs and reduce errors caused by human misjudgment.

[0056] According to the position information of the retrieval guide information in the prompt word template, the execution step to which the retrieval guide information belongs is determined, and the execution step is used as the target execution step. The retrieval guide information includes content information, and the content information indicates what kind of layout reference information is required for the target execution step. Therefore, the target description information adapted to the content information can be selected from the description information of the target space object according to the content information included in the retrieval guide information. The adaptation of the content information to the target description information means that the layout reference information indicated by the content information can be retrieved from the target layout knowledge base according to the target description information.

[0057] For example, if the content information included in the retrieval guide information is item type information, then the spatial type information of the target space object is obtained from the description information of the target space object as the target description information. For another example, if the content information included in the retrieval guide information is item size information, then the spatial size information and spatial type information of the target space object are obtained from the description information of the target space object as the target description information. For example, one implementation method of the retrieval guide information is: ${category_select_map}, wherein category_select_map(type_select_information) is the content information of the retrieval guide information, and ${} is the placeholder information of the retrieval guide information, i.e., an identifier. For ease of distinction and description, the placeholder information included in the retrieval guide information is referred to as the first placeholder information. Optionally, category_select_map may be used to replace the content information of the retrieval guide information with layout reference information to obtain layout reference information, i.e., "${layout reference information}". For a detailed introduction, please refer to the subsequent embodiments, which will not be repeated here.

[0058] Further optionally, the description information of the target space object includes: space attribute information and item layout type information. Wherein, the space attribute information refers to the attribute information corresponding to the target space object, for example, the space attribute information includes but is not limited to: space type information, space size information, and the location information of the door and window. The item layout type information refers to the item layout type information in the space object, for example, a straight line layout, a symmetrical layout, a compound layout, a central layout, an L-shaped layout, a U-shaped layout, etc. Wherein, the straight line layout means that the items in the target space object are arranged in a straight line, leaving space for activities around; the L-shaped layout means that the items in the target space object are arranged in an L-shape, which can improve the space utilization rate; the U-shaped layout means that the items in the target space object are arranged in a U-shape, which is suitable for space objects such as kitchens or operation rooms that require efficient operation; the symmetrical layout means that the space layout is centered on the central axis, and the two sides are symmetrically distributed, which is suitable for formal occasions or traditional style space objects; the central layout means that the target items are arranged around a certain center point or area as the core, which is suitable for places where concentrated activities are required, such as conference rooms, living rooms, etc. Wherein, different item layout type information corresponds to different layout rule knowledge information.

[0059] An implementation method of retrieving item attribute information or layout rule knowledge information adapted to target description information in a target layout knowledge base as layout reference information relied on by a target execution step includes: if the target execution step is a step for determining item attribute information, the step corresponds to an item attribute knowledge base, and the target description information is spatial attribute information of a target space object, then selecting item attribute information adapted to the spatial attribute information from the item attribute knowledge base as the item attribute information relied on by the step for determining item attributes; if the target execution step is a step for determining layout rule knowledge information, the step corresponds to a layout rule knowledge base, and the target description information is item layout type information, then selecting layout rule knowledge information adapted to the item layout type information from the layout rule knowledge base as the layout rule knowledge information relied on by the step for determining layout rule knowledge information. Among them, by comparing the spatial attribute information of the target space object with the data in the item attribute knowledge base, the item attribute information most suitable for the current scene can be accurately selected, so that the selection and configuration of items are more reasonable and applicable. By analyzing the item layout type information and selecting the corresponding layout rules, it is helpful to scientifically and rationally plan the spatial layout, maximize the space utilization rate, and ensure beauty and harmony at the same time.

[0060] Further optionally, the step for determining the item attribute information includes: a first step for determining the item type information, a second step for determining the item size information, and a fifth step for determining the item orientation information; the step for determining the layout rule knowledge information includes: a third step for determining the item layout description information and a fourth step for determining the item position information.

[0061] Among them, the first step corresponds to an item type knowledge base, and the target description information is the space type information of the target space object. Then, the item type information that is compatible with the space type information is selected from the item type knowledge base as the item type information that the first step depends on.

[0062] Among them, the second step corresponds to an object size knowledge base, and the target description information is the space type information and space size information of the target space object. Then, the object size information that is compatible with the space type information and space size information is selected from the object size knowledge base as the object size information that the second step depends on.

[0063] Among them, the third step corresponds to a layout description knowledge base, and the target description information is the layout type information of the target space object. Then, the object layout description information that is compatible with the layout type information is selected from the layout description knowledge base as the object layout description information that the third step depends on.

[0064] Among them, the fourth step corresponds to a spatial computing knowledge base, and the target description information is the spatial type information of the target spatial object. Then, spatial computing rule information adapted to the spatial type information is selected from the spatial computing knowledge base as the spatial computing rule knowledge information that the fourth step depends on.

[0065] Among them, the fifth step corresponds to an object orientation knowledge base, and the target description information is the layout type information of the target space object. Then, the object orientation information that is compatible with the layout type information is selected from the object orientation knowledge base as the object orientation information of the fifth step.

[0066] In an optional embodiment, the implementation method of calling the spatial layout generation model is not limited. An implementation method of calling the spatial layout generation model includes: generating target structured prompt words according to layout reference information and the initial spatial layout generation process; calling the spatial layout generation model according to the target structured prompt words. Among them, a prompt word generation method based on RAG technology and COT technology is provided, which uses RAG technology to retrieve layout reference information from a layout knowledge base, and uses COT technology to design the initial spatial layout generation process. The target structured prompt words can be dynamically generated, and the input information of the spatial layout generation model can be flexibly adjusted according to different needs and changes, so that the model output is more diversified and targeted, and meets various complex layout design needs.

[0067] The implementation method of calling the spatial layout generation model according to the target structured prompt word is not limited. For example, the target structured prompt word and the description information of the target spatial object are input into the spatial layout generation model to call the spatial layout generation model. For another example, the description information of the target spatial object is filled into the blank prompt word of the target structured prompt word to obtain a new target structured prompt word, and the new target structured prompt word is input into the spatial layout generation model to call the spatial layout generation model.

[0068] Optionally, an implementation method of generating target structured prompt words based on layout reference information and an initial space layout generation process includes: generating a target space layout generation process based on the layout reference information that at least some of the execution steps depend on, and the first placeholder information of the retrieval guide information included in at least some of the execution steps; generating target structured prompt words based on the target space layout generation process. For example, the target structured prompt words include the target space layout generation process, and further include description information of the target space object. Among them, the first placeholder information provides a structured framework for the execution step to which it belongs, clarifies the key information points that need to be filled in or considered in the execution step, helps to ensure the integrity and consistency of the execution steps, avoids missing important information, and makes the entire space layout process more systematic and orderly.

[0069] Wherein, the retrieval guide information includes the first placeholder information and the content information, for example, the retrieval guide information is category_select_map: ${}, wherein category_select_map is the content information, and ${} is the first placeholder information; based on this, the first placeholder information of the retrieval guide information included in at least some of the execution steps can be replaced with the layout reference information on which at least some of the execution steps depend, and the replaced initial space layout generation process is used as the target space layout generation process. For example, after replacing the first placeholder information, "category_select_map: layout reference information" is obtained. Alternatively, the retrieval guide information includes the first placeholder information, and the first placeholder information includes the content information, for example, ${category_select_map}, ${} is the first placeholder information, and category_select_map is the content information, then the layout reference information on which at least some of the execution steps depend is filled into the first placeholder information of the retrieval guide information included in at least some of the execution steps, and the filled initial space layout generation process is used as the target space layout generation process. For example, after filling the first placeholder information, "${layout reference information}" is obtained.

[0070] Optionally, in addition to the initial space layout generation process, the structured prompt word template also includes: initial task description information, wherein the initial task description information is used to describe the role played by the space layout generation model and the goals to be completed. The initial task description information includes placeholder information. For the convenience of distinction and description, the placeholder information included in the initial task description information is referred to as the second placeholder information; the initial space layout generation process includes: a step for determining layout rule knowledge information, which includes third placeholder information; the description information of the target space object includes: in addition to the space attribute information, the personalized layout description information input by the user. Among them, the personalized layout description information is the layout description information related to the placement of objects provided by the user according to his own needs. For example, "the living room does not need a coffee table and a sofa, but a long desk and chairs", "the bedroom needs to install a projector", etc.

[0071] Based on the above, an implementation method of generating target structured prompt words according to layout reference information and the initial space layout generation process includes: generating target task description information according to space attribute information and the second placeholder information; updating the target space layout generation process according to the personalized layout description information and the third placeholder information; generating target structured prompt words according to the target task description information and the updated target space layout generation process. For example, the target structured prompt words include the target task description information and the updated target space layout generation process. Further, the target structured prompt words also include: description information of the target space object. The second placeholder information and the third placeholder information provide a structured framework for the entire space layout process, clarify the key information points that need to be filled in or considered, so that the entire space layout process has integrity and consistency, and makes the entire space layout process more systematic and orderly. Among them, the introduction of personalized layout description information allows users to guide the generation of space layout according to their own needs and obtain space layout information that better meets user needs.

[0072] The implementation method of generating the target task description information according to the spatial attribute information and the second placeholder information is not limited. For example, the spatial attribute information can be filled into the second placeholder information, or the second placeholder information can be replaced with the spatial attribute information. For the introduction of filling and replacing, please refer to the description of the first placeholder information, which will not be repeated here. The initial task description information after filling or replacing can be used as the target task description information.

[0073] Among them, the implementation method of updating the target space layout generation process according to the personalized layout description information and the third placeholder information is not limited. For example, the personalized layout description information is filled into the third placeholder information, or the third placeholder information is replaced with the personalized layout description information, which is not limited to this. For the introduction of filling and replacing, please refer to the description of the first placeholder information, which will not be repeated here. The target space layout generation process after filling or replacement can be used as the updated target task description information. An example of a target structured prompt word is provided below, but it is not limited to this.

[0074] The target space layout generation process includes 6 steps, among which the first to the fifth steps rely on the layout reference information. Figure 2a As shown, an exemplary display is made for the target structured prompt words and the layout knowledge base, but it is not limited to this.

[0075] #Task description: You are a layout designer. For example, you can consider the room layout from a two-dimensional top-down perspective. The input target space object is a rectangular room, including space size information ${room_size}, space type information ${room_type} (such as living room or bedroom, etc.), door location information ${door_location}, and window location information ${windows_location}. Establish a coordinate system with the lower left corner of the target space object as the origin. Among them, "${room_size}", "${room_type}", "${door_location}" and "${windows_location}" are examples of the second placeholder information, but are not limited to this. Please place items that are suitable for the space type information in the target space object, such as furniture. The actual size information of the items can be selected according to the space size information, and these items are placed in the appropriate position of the target space object. For example, during the layout process, the rationality of the placement of the rectangular box of the object under the two-dimensional top-down perspective is considered, and finally all the objects are placed, and the spatial layout information of the objects is returned. The layout information includes the item type (category), item location information (such as the center point coordinates of the furniture center_x, center_y), item size information (length, width) and item orientation information (direction_angle).

[0076] #Initial space layout generation process:

[0077] Step 1: Select the appropriate item type (category) according to the space type information. Different space types correspond to different item type sets. The following is the item type information retrieved from the item type knowledge base according to the space type information: ${category_select_map}, where "${category_select_map}" is an example of the first placeholder information. You need to select the item type information (category) that is compatible with the target space object from the ${category_select_map} list. The following is the target item type information: ${category_ls}.

[0078] Step 2: According to the item type information selected by the spatial size information, set a length and width (length, width) that adapts to the spatial size information. The selection of the item length and width is related to the spatial size information. The following is the item size information ${size_select_map} retrieved from the item size knowledge base based on the spatial size information and the item type information, where "${size_select_map}" is an example of the first placeholder information, but is not limited to this. The item size information provides the size range corresponding to the selected item type information, and you need to select the target item size information of each item type information.

[0079] The third step: according to the item type information and item size information, combined with the item layout description information ${layout_reference} provided by the layout description knowledge base, determine the target item layout description information for item placement. Among them, "${layout_reference}" is an example of the first placeholder information, but is not limited to this. Further, in the process of generating the target item layout description information, personalized layout description information ${Personalized_layout_desc} can also be combined to meet the personalized needs of users. Among them, "${Personalized_layout_desc}" is an example of the third placeholder information, but is not limited to this. For example, consider how the items are placed in the space from a two-dimensional top-down perspective, and the target item layout description information can describe the location information of each item. For example, the position of the main furniture and the wall furniture can be described first, such as in the middle left of the room, close to the main wall, and then the position of the auxiliary furniture and matching furniture can be described, such as on the left side of the main furniture, 50cm away from the main furniture.

[0080] Step 4: Calculate the target item position information (e.g., two-dimensional geometric position) of each item according to the target item layout description information in step 3. The target item position information may be the center point position of the item, such as (center_x, center_y). The calculation process may refer to the spatial computing rule knowledge information. The following is the spatial computing rule knowledge information ${spatial_rules} retrieved from the spatial computing knowledge base, where “${spatial_rules}” is an example of the first placeholder information, but is not limited thereto.

[0081] Step 5: Calculate the target object orientation information of each object according to the target object layout description information in step 3. The orientation information of the object (direction_angle) indicates the orientation angle of the positive direction of the object, in degrees. For the establishment of the orientation information, please refer to the object orientation information ${object_direction_rules} retrieved from the object orientation knowledge base. Among them, "${object_direction_rules}" is an example of the first placeholder information, but is not limited to this.

[0082] Step 6: Return the target item type information, target item size information, target item position information and target item orientation information of each item in a fixed format. For example, spatial layout information = [(first furniture type, first furniture size, first furniture position, first furniture orientation), (second furniture type, second furniture size, second furniture position, second furniture orientation), ..., (Nth furniture type, Nth furniture size, Nth furniture position, Nth furniture orientation)], where N is a positive integer and N≥1.

[0083] In an optional embodiment, an implementation method of executing execution steps in a target space layout generation process according to layout reference information to obtain spatial layout information of a target space object includes: executing execution steps in a target space layout generation process according to spatial attribute information and layout reference information of the target space object to obtain intermediate reasoning information; determining the target objects and the placement relationship between the target objects as the spatial layout information of the target space object according to the intermediate reasoning information. Among them, intermediate reasoning information can be obtained after executing each execution step, and the intermediate reasoning information can be used as input information for the next execution step. By comprehensively considering the spatial attribute information and layout reference information of the target space object, the intermediate reasoning information is generated, so that each execution step has a sufficient basis, so that the generated spatial layout information is more accurate and reasonable, can better meet actual needs, and reduce the possibility of design errors and rework.

[0084] Optionally, the spatial attribute information of the target spatial object includes: spatial type information and spatial size information; the layout reference information includes: item type information, item size information, item layout description information, spatial calculation rule knowledge information and item orientation information; the target spatial layout generation process includes the first step to the fifth step;

[0085] A method for executing the execution steps in the target space layout generation process according to the spatial attribute information and layout reference information of the target space object to obtain the intermediate reasoning information includes: executing the first step according to the spatial type information of the target space object and the item type information to obtain the target item type information; executing the second step according to the spatial size information of the target space object and the target item type information and the item size information to obtain the target item size information; executing the third step according to the target item type information and the target item size information and the item layout description information to obtain the target item layout description information; executing the fourth step according to the target item layout description information and the spatial calculation rule knowledge information to obtain the target item position information; executing the fifth step according to the target item layout description information and the target item position information and the item orientation information to obtain the target item orientation information. Among them, through step-by-step, multi-level reasoning and calculation, each execution step has a clear basis and logic, so that the spatial layout information finally generated is more accurate and reasonable, can better meet the actual needs, and reduce the possibility of design errors and rework. In addition, through automated and systematic processes, labor costs are reduced and the generation efficiency and quality of spatial layout information are improved.

[0086] Further optionally, an implementation method of determining the target objects and the placement position relationship between the target objects as the spatial layout information of the target space object based on the intermediate reasoning information includes: determining the target objects suitable for placement in the target space object based on the target object type information and the target object size information; determining the placement position relationship between the target objects based on the target object position information and the target object orientation information, so as to obtain the spatial layout information corresponding to the target space object. By comprehensively considering the type and size information of the target objects, the selected objects are suitable for the target space object, so that the spatial layout information finally generated is more accurate and reasonable, can better meet the actual needs, and reduce the possibility of design errors and rework.

[0087] In an optional embodiment, in order to further improve the layout effect of the spatial layout information, the spatial layout information is optimized. For example, the optimization can be achieved through the following layout principles: layout principle E1: make sure that the items included in the spatial layout information do not overlap (i.e., collide); layout principle E2: make sure that the items included in the spatial layout information are all inside the target space object; layout principle E3: place some items against the wall, etc. One or more principles can be selected to optimize the spatial layout information to obtain the optimized spatial layout information.

[0088] Based on this, the spatial layout information generation method provided by the embodiment of the present application also includes: optimizing the position information and / or orientation information of each target object based on at least one layout principle in which the information of each target object in the spatial layout information does not overlap, each target object is located inside the target space object, and the distance between at least some target objects and the wall of the target space object is less than a set threshold, to obtain optimized spatial layout information. The layout principle can be flexibly selected according to different space requirements and object characteristics, which makes the design more flexible and adaptable, and can quickly respond to changes in user needs.

[0089] Among them, the optimization process can be a layout fine-tuning based on combinatorial optimization, wherein the layout fine-tuning refers to modeling the three-dimensional space layout information as a combinatorial optimization problem, and its objective function is to continuously adjust the position of the target object during the optimization process to minimize the value of some constraint models about the layout. Among them, the optimization process can make each target object in the target space object follow some spatial constraints, and the placement position is more reasonable, which meets the layout requirements of the user.

[0090] Among them, the independent variables are: represents the spatial location information of item i (including position and orientation), where x i =(x i ,y i ,z i ). Objective function: Where M is the number of constraint models (1≤M≤3), n is the number of target items included in the spatial layout information, and L m represents the loss function of the mth constraint model. The optimization goal is to minimize the global loss F of each constraint model. Each constraint model corresponds to a layout principle. The following introduces three layout principles.

[0091] Layout principle E1: Collision constraint, that is, the information of each target item does not overlap, that is, no physical penetration occurs. This ensures that there is no overlap between target items, avoiding space waste and inconvenience in use. The resulting spatial layout is more reasonable, items do not interfere with each other, and space utilization efficiency is improved.

[0092] For any two items i and j (i≠j), the collision constraint loss is defined as:

[0093] L collision =max(0,r i +r j -||x i -x j ||2)

[0094] where ||·||2 represents the Euclidean distance, r i and rj are the maximum practical radii of objects i and j respectively.

[0095] Layout principle E2: spatial coordinate constraint, that is, each target object is located inside the target space object. Ensure that all target objects are located inside the target space object to avoid objects exceeding the boundary and affecting the visual effect, thereby improving the user experience.

[0096] Assume that the boundary of the target space object is composed of the minimum point and maximum point Defined, the spatial coordinate constraint loss can be expressed as:

[0097]

[0098] Among them, x ik is the coordinate value of item i on axis k, ensuring that each dimension does not exceed the room boundary.

[0099] Layout principle E3: Wall constraint, that is, the distance between at least some of the target objects and the wall of the target space object is less than the set threshold. By adjusting the distance between the target objects and the wall, the spatial layout is more coordinated and beautiful. Reasonable spacing can avoid the space from appearing crowded or empty, creating a more harmonious visual effect.

[0100] Assuming n is the normal vector of the wall, for item i, the change in its distance to the wall can be expressed as:

[0101] Δd i =|n·(x i ′ -x i )|-d old

[0102] Among them, x i ′ is the new position of the item, d old is the distance from the original position of the object to the wall. The wall constraint loss can be expressed as: L wall-sticking =max(0,|Δd i |-∈), where ∈ is a small tolerance value that allows for slight position adjustments.

[0103] In summary, by integrating the loss functions of the above layout principles (constraint models), a global optimization problem can be constructed, with the goal of minimizing the overall loss of all constraint models, and optimization processing can be performed to achieve reasonable planning of spatial layout and obtain the final spatial layout information.

[0104] In an optional embodiment, the layout reference information may include: ground layout reference information and wall layout reference information; the space layout generation model includes: a ground layout generation model and a wall layout generation model; the ground layout generation model corresponds to a ground layout generation process, and the wall layout generation model corresponds to a wall layout generation process. The ground layout reference information includes layout reference information corresponding to items located on the ground, and the ground furniture may include but is not limited to: sofas, coffee tables, tables, chairs, beds, cabinets, and beds. The wall layout reference information includes layout reference information corresponding to items located on the wall, and the wall furniture may include but is not limited to: chandeliers, hanging paintings, projectors, televisions, etc. The ground layout reference information includes ground item attribute information or ground layout rule knowledge information as a constraint condition for executing the steps, and the wall layout reference information includes wall item attribute information or wall layout rule knowledge information as a constraint condition for executing the steps, wherein the detailed implementation methods of the ground layout reference information and the wall layout reference information can refer to the aforementioned introduction to the layout reference information, which will not be repeated here.

[0105] Optionally, according to the description information of the target space object, retrieving from a layout knowledge base at least some of the layout reference information that the execution steps depend on, comprises:

[0106] According to the ground description information of the target space object, at least some of the ground layout reference information on which each execution step depends is retrieved from the layout knowledge base, and according to the wall description information of the target space object, at least some of the wall layout reference information on which each execution step depends is retrieved from the layout knowledge base. The ground description information includes the ground attribute information and ground layout type information of the target space object, and the wall description information includes the wall attribute information and wall layout type information of the target space object. For a detailed introduction, please refer to the aforementioned space attribute information and space layout type information, which will not be repeated here.

[0107] Optionally, a method of calling a spatial layout generation model to execute the execution steps in the target spatial layout generation process according to the layout reference information to obtain the spatial layout information of the target spatial object includes: calling a ground layout generation model to execute the execution steps in the ground layout generation process according to the ground layout reference information to obtain the ground layout information of the target spatial object; calling a wall layout generation model to execute the execution steps in the wall layout generation process according to the wall layout reference information and the ground layout information to obtain the wall layout information of the target spatial object; obtaining the spatial layout information of the target spatial object according to the ground layout information and the wall layout information. By jointly generating spatial layout information with the ground layout generation model and the wall layout generation model, the comprehensiveness and completeness of the spatial layout can be improved. Effect: The spatial layout information finally generated covers all the details of the ground and the wall, making the design plan more complete and detailed. Among them, for a detailed introduction to the above-mentioned method, please refer to the aforementioned implementation method of calling the spatial layout generation model, which will not be repeated here.

[0108] In an optional embodiment, the embodiment of the present application provides Figure 1 The spatial layout information generation method shown can be implemented by a spatial layout agent, which is a software entity that uses advanced natural language processing technology, especially a large pre-trained language model, to generate spatial layout information. The agent completes the spatial layout generation task by understanding layout knowledge and natural language (e.g., description information of the target spatial object), and its core lies in the large language model it includes.

[0109] Among them, the spatial layout agent can automatically analyze and generate spatial layout information based on layout reference information and user needs, reduce the time and cost of manual intervention, and improve the flexibility and adaptability of the layout.

[0110] Optionally, the spatial layout agent includes: a ground layout generation agent and a wall layout generation agent. The ground layout generation agent is used to obtain ground description information of the target space object for which spatial layout information is to be generated; obtain a ground layout generation process corresponding to the ground layout generation model, at least part of the execution steps in the ground layout generation process depend on the ground layout reference information, and different execution steps depend on different ground layout reference information; according to the ground description information of the target space object, retrieve the ground layout reference information that at least part of the execution steps depend on from the ground layout knowledge base, the ground layout reference information includes ground object attribute information or ground layout rule knowledge information as a constraint condition for the execution step; call the ground layout generation model, and execute the execution steps in the ground layout generation process according to the ground layout reference information to obtain the ground space layout information of the target space object; wherein the ground space layout information is used to describe the target objects suitable for being placed on the ground of the target space object and the placement relationship between the target objects.

[0111] The wall layout generation agent is used to obtain the wall description information and ground space layout information of the target space object for which the spatial layout information is to be generated; obtain the wall layout generation process corresponding to the wall layout generation model, at least part of the execution steps in the wall layout generation process depend on the wall layout reference information, and the wall layout reference information that different execution steps depend on is different; according to the wall description information of the target space object, retrieve the wall layout reference information that at least part of the execution steps depend on from the wall layout knowledge base, the wall layout reference information includes the wall object attribute information or the wall layout rule knowledge information as the constraint condition of the execution step; call the wall layout generation model, and execute the execution steps in the wall layout generation process according to the wall layout reference information and the ground space layout information to obtain the spatial layout information of the target space object; wherein the spatial layout information is used to describe the target objects suitable for being placed in the target space object and the placement position relationship between the target objects.

[0112] Among them, the ground layout generation agent is responsible for generating ground layout information, which can optimize the utilization and functionality of ground space more finely. The wall layout generation agent is responsible for generating wall layout information, which can more accurately consider the aesthetics and practicality of the wall. Each agent only needs to handle specific tasks, reducing the computational burden of a single agent and improving processing speed.

[0113] Among them, the detailed implementation methods of the above-mentioned space layout intelligent agent, ground layout generation intelligent agent and wall layout generation intelligent agent can be found in the above-mentioned embodiments, which will not be repeated here.

[0114] like Figure 2bAs shown, the spatial layout information generation process is exemplarily demonstrated. The spatial layout information generation process is divided into three layers, namely, the input layer, the spatial layout generation layer and the output layer.

[0115] The input layer mainly implements the following functions: obtaining the description information of the target space object; performing normalization processing on the target space object; obtaining the space type information, space size information, and door and window position information for the target space object after normalization processing. Normalization processing refers to determining the main wall of the target space object, and establishing the spatial coordinate system of the target space object according to the position of the main wall, wherein the main wall can be the wall that occupies the visual main body in the target space object, for example, the main wall can be a wall that does not include a door or a window. Based on the spatial coordinate system, the space size information and door and window position information are obtained.

[0116] Among them, the spatial layout generation layer mainly realizes the following functions: retrieving layout reference information from the layout knowledge base, for example, retrieving ground layout reference information from the layout knowledge base according to the ground description information of the target spatial object; adding the ground layout reference information to the ground structured prompt word template to generate ground layout prompt words; retrieving wall layout reference information from the layout knowledge base according to the wall description information of the target spatial object; adding the wall layout reference information to the wall structured prompt word template to generate wall layout prompt words; inputting the ground layout prompt words into the ground layout generation model, executing the execution steps in the ground layout generation process in the ground layout prompt words to obtain the ground layout information of the target spatial object; inputting the ground layout information and the wall layout prompt words into the wall layout generation model, executing the execution steps in the wall layout generation process in the wall layout prompt words to obtain the spatial layout information of the target spatial object.

[0117] The output layer mainly implements the following functions: the spatial layout information includes the attribute information of each target item; according to the attribute information of each target item, each target item is mapped to the spatial coordinate system of the target space object to realize the coordinate translation transformation. The transformed target space object and each target item included therein are optimized, for example, by optimizing at least one of the above layout principles E1 to E3, to obtain the optimized spatial layout information.

[0118] In an optional embodiment, the plane layout information of the target space object in a two-dimensional top-down perspective can be generated based on the space layout information; further, the target object can be obtained from the object library according to the size information of the target object, and three-dimensional rendering is performed on each target object in the target space object to obtain the three-dimensional layout information of the target space object. Taking the target space object as a bedroom as an example, Figure 3a and Figure 3bThe plane layout information and three-dimensional layout information are shown in an exemplary manner. The arrows indicate the orientation of the objects. For example, the target space object is a bedroom. Figure 4a and Figure 4b Another exemplary display of planar layout information and three-dimensional layout information. Taking the target space object as a living room as an example, Figure 5a and Figure 5b Another exemplary embodiment shows planar layout information and three-dimensional layout information respectively.

[0119] Regarding the embodiments of the present application Figure 1 The detailed implementation and beneficial effects of each step in the method have been described in detail in the aforementioned embodiments and will not be elaborated here.

[0120] It should be noted that the execution subject of each step of the method provided in the above embodiment can be the same device, or the method can be executed by different devices. For example, the execution subject of steps 101 to 103 can be device A; for another example, the execution subject of steps 101 and 102 can be device A, and the execution subject of step 103 can be device B; and so on.

[0121] In addition, in some of the processes described in the above embodiments and the accompanying drawings, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel, and the sequence numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the sequence numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0122] Figure 6 A schematic diagram of a structure of a spatial layout information generating device provided by an exemplary embodiment of the present application is shown in FIG. Figure 6 As shown, the device includes: an acquisition module 61, a retrieval module 62 and a processing module 63.

[0123] An acquisition module is used to acquire description information of a target space object for which spatial layout information is to be generated;

[0124] An acquisition module, used for acquiring a target space layout generation process corresponding to the space layout generation model, wherein at least some execution steps in the target space layout generation process depend on the layout reference information, and different execution steps depend on different layout reference information;

[0125] A retrieval module, configured to retrieve, from a layout knowledge base, at least some of the layout reference information that the execution steps depend on, based on the description information of the target space object, wherein the layout reference information includes item attribute information or layout rule knowledge information that is a constraint condition for the execution steps;

[0126] A processing module is used to call the spatial layout generation model and execute the execution steps in the target spatial layout generation process according to the layout reference information to obtain the spatial layout information of the target spatial object;

[0127] The spatial layout information is used to describe target objects that are suitable for placement in the target space object and the placement relationship between the target objects.

[0128] In an optional embodiment, the spatial layout generation model corresponds to a structured prompt word template, the structured prompt word template includes an initial spatial layout generation process, and at least some of the execution steps in the initial spatial layout generation process include retrieval guidance information; the retrieval module is specifically used to: under the guidance of the retrieval guidance information, according to the description information of the target spatial object, retrieve from the layout knowledge base at least some of the layout reference information that the execution steps depend on.

[0129] Optionally, the retrieval module is specifically used to: select a target layout knowledge base that is adapted to the target execution step information from multiple layout knowledge bases based on the target execution step information belonging to the retrieval guidance information, and the target execution step belongs to at least part of the execution step; select target description information that is adapted to the content information from the description information based on the content information of the retrieval guidance information; retrieve object attribute information or layout rule knowledge information that is adapted to the target description information in the target layout knowledge base as layout reference information that the target execution step depends on.

[0130] Further optionally, the description information of the target space object includes: spatial attribute information and item layout type information; the retrieval module is specifically used for: if the target execution step is a step for determining item attribute information, the step corresponds to an item attribute knowledge base, and the target description information is the spatial attribute information of the target space object, then the item attribute information that is compatible with the spatial attribute information is selected from the item attribute knowledge base as the item attribute information that the step for determining the item attributes depends on; if the target execution step is a step for determining layout rule knowledge information, the step corresponds to a layout rule knowledge base, and the target description information is item layout type information, then the layout rule knowledge information that is compatible with the item layout type information is selected from the layout rule knowledge base as the layout rule knowledge information that the step for determining the layout rule knowledge information depends on.

[0131] Optionally, the processing module is specifically used to: generate target structured prompt words according to the layout reference information and the initial space layout generation process; and call the space layout generation model according to the target structured prompt words.

[0132] Further optionally, the retrieval guide information includes the first placeholder information, and the processing module is specifically used to: generate a target space layout generation process based on the layout reference information on which at least some of the execution steps depend, and the first placeholder information of the retrieval guide information included in at least some of the execution steps; generate a target structured prompt word based on the target space layout generation process.

[0133] Further optionally, the structured prompt word template also includes: initial task description information, the initial task description information includes second placeholder information; the target space layout generation process includes steps for determining layout rule knowledge information, the steps include third placeholder information; the description information of the target space object includes: space attribute information and personalized layout description information; the processing module is specifically used to: generate target task description information based on the space attribute information and the second placeholder information; update the target space layout generation process based on the personalized layout description information and the third placeholder information; generate target structured prompt words based on the target task description information and the updated target space layout generation process.

[0134] In an optional embodiment, the processing module is specifically used to: execute the execution steps in the target space layout generation process according to the spatial attribute information and layout reference information of the target space object to obtain intermediate reasoning information; and determine the target objects and the placement position relationship between the target objects as the spatial layout information of the target space object according to the intermediate reasoning information.

[0135] Optionally, the spatial attribute information of the target space object includes spatial type information and spatial size information; the layout reference information includes: item type information, item size information, item layout description information, spatial calculation rule knowledge information and item orientation information; the target space layout generation process includes the first step to the fifth step; the processing module is specifically used to: execute the first step according to the spatial type information of the target space object, combined with the item type information, to obtain the target item type information; execute the second step according to the spatial size information of the target space object and the target item type information, combined with the item size information, to obtain the target item size information; execute the third step according to the target item type information and the target item size information, combined with the item layout description information, to obtain the target item layout description information; execute the fourth step according to the target item layout description information, combined with the spatial calculation rule knowledge information, to obtain the target item position information; execute the fifth step according to the target item layout description information and the target item position information, combined with the item orientation information, to obtain the target item orientation information.

[0136] Optionally, the processing module is specifically used to: determine the target objects suitable for placement in the target space object based on the target object type information and the target object size information; determine the placement position relationship between the target objects based on the target object position information and the target object orientation information, so as to obtain the spatial layout information corresponding to the target space object.

[0137] In an optional embodiment, the device further includes an optimization module. The optimization module is used to optimize the position information and / or orientation information of each target object based on at least one layout principle of no overlap between target object information in the spatial layout information, each target object being located inside the target spatial object, and the distance between at least some of the target objects and the wall of the target spatial object being less than a set threshold, to obtain optimized spatial layout information.

[0138] In an optional embodiment, the layout reference information includes: ground layout reference information and wall layout reference information; the spatial layout generation model includes a ground layout generation model and a wall layout generation model; the ground layout generation model corresponds to a ground layout generation process, and the wall layout generation model corresponds to a wall layout generation process; the processing module is specifically used to: call the ground layout generation model, and execute the execution steps in the ground layout generation process according to the ground layout reference information to obtain the ground layout information of the target space object; call the wall layout generation model, and execute the execution steps in the wall layout generation process according to the wall layout reference information and the ground layout information to obtain the wall layout information of the target space object; obtain the spatial layout information of the target space object according to the ground layout information and the wall layout information.

[0139] Regarding the embodiments of the present application Figure 6 The detailed implementation and beneficial effects of each step in the illustrated device have been described in detail in the aforementioned embodiments and will not be elaborated on here.

[0140] Figure 7 A schematic diagram of the structure of an electronic device provided by an exemplary embodiment of the present application is shown in FIG. Figure 7 As shown, the device includes: a memory 74 and a processor 75.

[0141] The memory 74 is used to store computer programs and can be configured to store various other data to support operations on the electronic device. Examples of such data include instructions for any application or method used to operate on the electronic device.

[0142] The processor 75 is coupled to the memory 74 and is used to execute the computer program in the memory 74, so as to: obtain the description information of the target space object for which the spatial layout information is to be generated; obtain the target space layout generation process corresponding to the spatial layout generation model, wherein at least part of the execution steps in the target space layout generation process depend on the layout reference information, and different execution steps depend on different layout reference information; according to the description information of the target space object, retrieve the layout reference information on which at least part of the execution steps depend from the layout knowledge base, wherein the layout reference information includes the item attribute information or layout rule knowledge information as the constraint condition of the execution step; call the spatial layout generation model, and execute the execution steps in the target space layout generation process according to the layout reference information to obtain the spatial layout information of the target space object; wherein the spatial layout information is used to describe the target items suitable for being placed in the target space object and the placement position relationship between the target items.

[0143] In an optional embodiment, the spatial layout generation model corresponds to a structured prompt word template, the structured prompt word template includes an initial spatial layout generation process, and at least some of the execution steps in the initial spatial layout generation process include retrieval guidance information; when the processor 75 retrieves the layout reference information on which at least some of the execution steps depend from the layout knowledge base according to the description information of the target spatial object, it is specifically used to: under the guidance of the retrieval guidance information, retrieve at least some of the layout reference information on which the execution steps depend from the layout knowledge base according to the description information of the target spatial object.

[0144] Optionally, under the guidance of the retrieval guidance information, the processor 75 retrieves the layout reference information on which at least some of the execution steps depend from the layout knowledge base according to the description information of the target space object, and is specifically used to: select a target layout knowledge base that is adapted to the target execution step information from multiple layout knowledge bases according to the target execution step information belonging to the retrieval guidance information, and the target execution step belongs to at least some of the execution steps; select target description information that is adapted to the content information from the description information according to the content information of the retrieval guidance information; and retrieve object attribute information or layout rule knowledge information that is adapted to the target description information from the target layout knowledge base as the layout reference information on which the target execution step depends.

[0145] Further optionally, the description information of the target space object includes: spatial attribute information and item layout type information; when the processor 75 retrieves the item attribute information or layout rule knowledge information that is compatible with the target description information from the target layout knowledge base as the layout reference information that the target execution step depends on, it is specifically used for: if the target execution step is a step for determining item attribute information, the step corresponds to an item attribute knowledge base, and the target description information is the spatial attribute information of the target space object, then the item attribute information that is compatible with the spatial attribute information is selected from the item attribute knowledge base as the item attribute information that the step for determining the item attributes depends on; if the target execution step is a step for determining layout rule knowledge information, the step corresponds to a layout rule knowledge base, and the target description information is item layout type information, then the layout rule knowledge information that is compatible with the item layout type information is selected from the layout rule knowledge base as the layout rule knowledge information that the step for determining the layout rule knowledge information depends on.

[0146] Optionally, when calling the spatial layout generation model, the processor 75 is specifically used to: generate target structured prompt words according to the layout reference information and the initial spatial layout generation process; and call the spatial layout generation model according to the target structured prompt words.

[0147] Further optionally, the retrieval guide information includes the first placeholder information, and when the processor 75 generates the target structured prompt words according to the layout reference information and the initial spatial layout generation process, it is specifically used to: generate a target spatial layout generation process according to the layout reference information on which at least some of the execution steps depend, and the first placeholder information of the retrieval guide information included in at least some of the execution steps; generate a target structured prompt word according to the target spatial layout generation process.

[0148] Further optionally, the structured prompt word template also includes: initial task description information, the initial task description information includes second placeholder information; the target space layout generation process includes steps for determining layout rule knowledge information, the steps include third placeholder information; the description information of the target space object includes: spatial attribute information and personalized layout description information; when the processor 75 generates the target structured prompt words according to the target space layout generation process, it is specifically used to: generate the target task description information according to the spatial attribute information and the second placeholder information; update the target space layout generation process according to the personalized layout description information and the third placeholder information; generate the target structured prompt words according to the target task description information and the updated target space layout generation process.

[0149] In an optional embodiment, when the processor 75 executes the execution steps in the target space layout generation process according to the layout reference information to obtain the spatial layout information of the target space object, it is specifically used to: execute the execution steps in the target space layout generation process according to the spatial attribute information and layout reference information of the target space object to obtain intermediate reasoning information; determine the target objects and the placement position relationship between the target objects as the spatial layout information of the target space object according to the intermediate reasoning information.

[0150] Optionally, the spatial attribute information of the target space object includes spatial type information and spatial size information; the layout reference information includes: item type information, item size information, item layout description information, spatial calculation rule knowledge information and item orientation information; the target space layout generation process includes the first step to the fifth step; when the processor 75 executes the execution steps in the target space layout generation process according to the spatial attribute information and layout reference information of the target space object to obtain the intermediate reasoning information, it is specifically used to: execute the first step according to the spatial type information of the target space object, combined with the item type information, to obtain the target item type information; execute the second step according to the spatial size information of the target space object and the target item type information, combined with the item size information, to obtain the target item size information; execute the third step according to the target item type information and the target item size information, combined with the item layout description information, to obtain the target item layout description information; execute the fourth step according to the target item layout description information, combined with the spatial calculation rule knowledge information, to obtain the target item position information; execute the fifth step according to the target item layout description information and the target item position information, combined with the item orientation information, to obtain the target item orientation information.

[0151] Optionally, when the processor 75 determines the target objects and the placement position relationships between the target objects as the spatial layout information of the target space object based on the intermediate reasoning information, it is specifically used to: determine the target objects suitable for being placed in the target space object based on the target object type information and the target object size information; determine the placement position relationships between the target objects based on the target object position information and the target object orientation information, so as to obtain the spatial layout information corresponding to the target space object.

[0152] In an optional embodiment, the processor 75 is further used to optimize the position information and / or orientation information of each target object based on at least one layout principle: there is no overlap between the target object information in the spatial layout information, each target object is located inside the target space object, and the distance between at least some of the target objects and the wall of the target space object is less than a set threshold, so as to obtain optimized spatial layout information.

[0153] In an optional embodiment, the layout reference information includes: ground layout reference information and wall layout reference information; the space layout generation model includes a ground layout generation model and a wall layout generation model; the ground layout generation model corresponds to a ground layout generation process, and the wall layout generation model corresponds to a wall layout generation process;

[0154] When the processor 75 calls the spatial layout generation model and executes the execution steps in the target spatial layout generation process according to the layout reference information to obtain the spatial layout information of the target spatial object, it is specifically used to: call the ground layout generation model and execute the execution steps in the ground layout generation process according to the ground layout reference information to obtain the ground layout information of the target spatial object; call the wall layout generation model and execute the execution steps in the wall layout generation process according to the wall layout reference information and the ground layout information to obtain the wall layout information of the target spatial object; obtain the spatial layout information of the target spatial object according to the ground layout information and the wall layout information.

[0155] Regarding the embodiments of the present application Figure 7 The detailed implementation and beneficial effects of each step in the illustrated device have been described in detail in the aforementioned embodiments and will not be elaborated on here.

[0156] Further, if Figure 7 As shown, the electronic device also includes: a communication component 76, a display 77, a power component 78, an audio component 79 and other components. Figure 7 Only some components are shown schematically, which does not mean that the electronic device only includes Figure 7 In addition, Figure 7 The components in the dashed box are optional components, not mandatory components, and the specific components depend on the product form of the electronic device. The electronic device of this embodiment can be implemented as a terminal device such as a desktop computer, a laptop computer, a smart phone, or an IOT (Internet of Things) device and a smart wearable device (such as a smart watch, a smart bracelet), or a conventional server, a cloud server, or a server array. If the electronic device of this embodiment is implemented as a terminal device such as a desktop computer, a laptop computer, a smart phone, etc., it can include Figure 7 If the electronic device of this embodiment is implemented as a server device such as a conventional server, a cloud server or a server array, it may not include Figure 7 Components within the dashed box.

[0157] The detailed implementation and beneficial effects of the electronic device provided in the embodiments of the present application have been described in detail in the aforementioned embodiments and will not be elaborated here.

[0158] Accordingly, the present application also provides a computer-readable storage medium storing a computer program, which can achieve the above-mentioned Figure 1 The steps in the method embodiment shown may be executed by an electronic device.

[0159] Accordingly, the embodiment of the present application also provides a computer program product, including a computer program / instruction, which, when executed by a processor, causes the processor to implement the above Figure 1 The steps in the method embodiment shown may be executed by an electronic device.

[0160] The above-mentioned memory can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read only memory (EEPROM), erasable programmable read only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0161] The above-mentioned communication component is configured to facilitate wired or wireless communication between the device where the communication component is located and other devices. The device where the communication component is located can access a wireless network based on a communication standard, such as WiFi, 2G, 3G, 4G / LTE, 5G and other mobile communication networks, or a combination thereof. In an exemplary embodiment, the communication component receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be based on Radio Frequency Identification (RFID) technology, Infrared Data Association (IrDA) technology, Ultra Wide Band (UWB) technology, Bluetooth (BT) technology and other technologies.

[0162] The above-mentioned display includes a screen, and the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation.

[0163] The power supply assembly provides power to various components of the device where the power supply assembly is located. The power supply assembly may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the device where the power supply assembly is located.

[0164] The above-mentioned audio component can be configured to output and / or input audio signals. For example, the audio component includes a microphone (Microphone, MIC), and when the device where the audio component is located is in an operating mode, such as a call mode, a recording mode, and a speech recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in a memory or sent via a communication component. In some embodiments, the audio component also includes a speaker for outputting an audio signal.

[0165] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-readable storage media (including but not limited to disk storage, compact disc read-only memory (Compact Disc Read-Only Memory, CD-ROM), optical storage, etc.) containing computer-usable program code.

[0166] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0167] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0168] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0169] In a typical configuration, a computing device includes one or more processors (Central Processing Unit, CPU), input / output interface, network interface and memory.

[0170] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0171] Computer readable media include permanent and non-permanent, removable and non-removable media that can be used to store information by any method or technology. Information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase-change random access memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0172] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0173] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for generating spatial layout information, characterized in that: include: Obtain description information of a target space object for which spatial layout information is to be generated; Acquire a target space layout generation process corresponding to the space layout generation model, wherein at least some execution steps in the target space layout generation process depend on layout reference information, and different execution steps depend on different layout reference information; According to the description information of the target space object, the layout reference information on which the at least part of the execution steps respectively depend is retrieved from a layout knowledge base, wherein the layout reference information includes item attribute information or layout rule knowledge information as a constraint condition of the execution step; Calling the spatial layout generation model, and executing the execution steps in the target spatial layout generation process according to the layout reference information, so as to obtain the spatial layout information of the target spatial object; The spatial layout information is used to describe target objects suitable for being placed in the target space object and the placement relationship between the target objects.

2. The method according to claim 1, characterized in that The spatial layout generation model corresponds to a structured prompt word template, the structured prompt word template includes an initial spatial layout generation process, and at least part of the execution steps in the initial spatial layout generation process includes retrieving guidance information; According to the description information of the target space object, retrieving the layout reference information on which the at least part of the execution steps respectively depend from the layout knowledge base, including: Under the guidance of the retrieval guidance information, the layout reference information on which the at least part of the execution steps respectively depend is retrieved from the layout knowledge base according to the description information of the target space object.

3. The method according to claim 2, characterized in that Under the guidance of the retrieval guidance information, according to the description information of the target space object, the layout reference information on which the at least part of the execution steps respectively depend is retrieved from the layout knowledge base, including: According to the target execution step information to which the search guide information belongs, selecting a target layout knowledge base adapted to the target execution step information from a plurality of layout knowledge bases, the target execution step belonging to the at least part of the execution steps; According to the content information of the search guide information, selecting target description information adapted to the content information from the description information; The object attribute information or layout rule knowledge information adapted to the target description information is retrieved from the target layout knowledge base as layout reference information that the target execution step depends on.

4. The method according to claim 3, characterized in that The description information of the target space object includes: space attribute information and item layout type information; Retrieving item attribute information or layout rule knowledge information that is compatible with the target description information in the target layout knowledge base as layout reference information that the target execution step depends on, including: If the target execution step is a step for determining item attribute information, the step corresponds to an item attribute knowledge base, and the target description information is spatial attribute information of a target spatial object, item attribute information that is compatible with the spatial attribute information is selected from the item attribute knowledge base as the item attribute information that the step for determining item attributes depends on; If the target execution step is a step for determining layout rule knowledge information, the step corresponds to a layout rule knowledge base, and the target description information is object layout type information, then layout rule knowledge information that is compatible with the object layout type information is selected from the layout rule knowledge base as the layout rule knowledge information that the step for determining the layout rule knowledge information depends on.

5. The method according to claim 2, characterized in that: Calling the spatial layout generation model includes: Generate target structured prompt words according to the layout reference information and the initial space layout generation process; The spatial layout generation model is called according to the target structured prompt word.

6. The method according to claim 5, characterized in that The first placeholder information included in the search guide information generates a target structured prompt word according to the layout reference information and the initial space layout generation process, including: Generate a target space layout generation process according to the layout reference information that the at least some of the execution steps respectively depend on, and the first placeholder information of the retrieval guide information respectively included in the at least some of the execution steps; According to the target space layout generation process, target structured prompt words are generated.

7. The method according to claim 6, characterized in that The structured prompt word template further includes: initial task description information, the initial task description information includes second placeholder information; the target space layout generation process includes a step for determining layout rule knowledge information, the step includes third placeholder information; the description information of the target space object includes: space attribute information and personalized layout description information; according to the target space layout generation process, generating a target structured prompt word includes: Generate target task description information according to the space attribute information and the second placeholder information; According to the personalized layout description information and the third placeholder information, updating the target space layout generation process; Generate target structured prompt words according to the target task description information and the updated target space layout generation process.

8. The method according to any one of claims 1 to 7, characterized in that: Executing the execution steps in the target space layout generation process according to the layout reference information to obtain the space layout information of the target space object includes: According to the spatial attribute information of the target spatial object and the layout reference information, executing the execution steps in the target spatial layout generation process to obtain intermediate reasoning information; According to the intermediate reasoning information, the target objects and the placement position relationships between the target objects are determined as the spatial layout information of the target space object.

9. The method according to claim 8, characterized in that The spatial attribute information of the target spatial object includes spatial type information and spatial size information; the layout reference information includes: item type information, item size information, item layout description information, spatial calculation rule knowledge information and item orientation information; the target spatial layout generation process includes the first step to the fifth step; According to the spatial attribute information of the target spatial object and the layout reference information, executing the execution steps in the target spatial layout generation process to obtain intermediate reasoning information includes: According to the space type information of the target space object and in combination with the item type information, the first step is performed to obtain the target item type information; According to the spatial dimension information of the target spatial object and the target item type information, combined with the item dimension information, a second step is performed to obtain the target item dimension information; According to the target item type information and the target item size information, combined with the item layout description information, a third step is performed to obtain the target item layout description information; According to the target object layout description information, combined with the spatial calculation rule knowledge information, the fourth step is performed to obtain the target object location information; According to the target object layout description information and the target object position information, combined with the object orientation information, the fifth step is performed to obtain the target object orientation information.

10. The method according to claim 9, characterized in that Determining the target objects and the placement relationship between the target objects as the spatial layout information of the target space object according to the intermediate reasoning information includes: Determining a target object suitable for being placed in the target space object according to the target object type information and the target object size information; According to the target object position information and the target object orientation information, the placement position relationship between the target objects is determined to obtain the spatial layout information corresponding to the target space object.

11. The method according to any one of claims 1-7, 9-10, characterized in that: The method further comprises: Based on at least one layout principle of no overlap between the target object information in the spatial layout information, the target objects being located inside the target spatial object, and the distance between at least some of the target objects and the wall of the target spatial object being less than a set threshold, the position information and / or orientation information of the target objects are optimized to obtain optimized spatial layout information.

12. The method according to any one of claims 1-7, 9-10, characterized in that: The layout reference information includes: ground layout reference information and wall layout reference information; the space layout generation model includes a ground layout generation model and a wall layout generation model; the ground layout generation model corresponds to a ground layout generation process, and the wall layout generation model corresponds to a wall layout generation process; Calling the spatial layout generation model and executing the execution steps in the target spatial layout generation process according to the layout reference information to obtain the spatial layout information of the target spatial object, including: Calling the ground layout generation model, and executing the execution steps in the ground layout generation process according to the ground layout reference information to obtain the ground layout information of the target space object; Calling the wall layout generation model, and executing the execution steps in the wall layout generation process according to the wall layout reference information and the ground layout information, so as to obtain the wall layout information of the target space object; The spatial layout information of the target spatial object is obtained according to the ground layout information and the wall layout information.

13. An electronic device, characterized in that: include: Memory and processor; The memory is used to store a computer program; the processor is coupled to the memory and is used to execute the computer program to implement the steps in the method according to any one of claims 1 to 12.

14. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the processor is caused to implement the steps in the method according to any one of claims 1 to 12.

15. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the processor is caused to implement the steps in the method according to any one of claims 1 to 12.

Images