Spatial layout method and spatial layout device

Abstract

The invention discloses a spatial layout method and a spatial layout device, and belongs to the field of spatial layout. The spatial layout method comprises the steps of inputting category information of a target space into a target model, and obtaining at least one component output by the target model and associated with the target space and initial pose information corresponding to the component; optimizing the initial pose information to obtain target pose information; and determining a layout result of the target space based on the target pose information. According to the spatial layout method, model training is performed by learning the placement relationship between the components and the relationship between the furniture and the room type, so that the complexity of the model can be reduced, a part of tasks can still be completed in a small amount of data, the convergence of the model is improved, and on the basis of ensuring the model training precision, the model training efficiency is improved. Requirements on data and network depth are remarkably reduced, operation efficiency is improved, and the method is suitable for complex and harsh layout scenes and has universality.

Description

Technical Field

[0001] The present application belongs to the field of space layout, and in particular, relates to a space layout method and a space layout device. Background Art

[0002] In the related art, when performing spatial layout, the layout is often performed by pre-setting layout rules. However, this method cannot be applied to complex and changeable spatial scenes, consumes a lot of manpower, material resources and time costs, has low layout efficiency, and has poor layout effects. Summary of the invention

[0003] The present application aims to solve at least one of the technical problems existing in the prior art. To this end, the present application proposes a spatial layout method and a spatial layout device, which significantly reduce the requirements for data and network depth on the basis of ensuring the accuracy of model training, improve the computing efficiency, and are applicable to complex and demanding layout scenarios and have universal applicability.

[0004] In a first aspect, the present application provides a spatial layout method, the method comprising: Inputting category information of the target space into the target model, obtaining at least one component associated with the target space output by the target model, and initial pose information corresponding to the component; Optimizing the initial posture information to obtain target posture information; Based on the target posture information, determining a layout result of the target space; The target model is obtained through training using the category information of the sample space and the sample layout information corresponding to the sample space as training samples.

[0005] According to the spatial layout method of the present application, by learning the placement relationship between components and the relationship between furniture and room types to perform model training, the complexity of the model can be reduced so that part of the task can still be completed with a small amount of data, the convergence of the model can be improved, and on the basis of ensuring the accuracy of model training, the requirements for data and network depth are significantly reduced, the computing efficiency is improved, and it is suitable for complex and demanding layout scenarios and has universal applicability.

[0006] According to an embodiment of the present application, optimizing the initial pose information to obtain target pose information includes: The at least one component is grouped to obtain at least one component set; the components in the same component set are associated with each other, and / or the components belong to the same category; While keeping the initial pose information between the components in the component set unchanged, the initial pose information between the component sets is optimized to obtain the target pose information corresponding to each component set.

[0007] According to one embodiment of the present application, grouping the at least one component includes: receiving a first input from a user, where the first input is used to input group information; In response to the first input, the at least one component is grouped based on the grouping information.

[0008] According to an embodiment of the present application, optimizing the initial pose information to obtain target pose information includes: A heuristic algorithm is used to optimize the initial pose information to obtain the target pose information.

[0009] According to an embodiment of the present application, optimizing the initial pose information to obtain target pose information includes: When the number of optimization times of optimizing the initial pose information exceeds an optimization threshold and an optimal solution is still not obtained, based on the area of ​​the target space and / or the size of the component, at least one component is deleted from all the components to obtain a target component; Optimize the initial pose information corresponding to the target component to obtain the target pose information.

[0010] According to one embodiment of the present application, the initial posture information includes at least one of a relative position and a rotation relationship.

[0011] In a second aspect, the present application provides a spatial layout device, the device comprising: A first processing module, configured to input category information of a target space into a target model, and obtain at least one component associated with the target space output by the target model, and initial pose information corresponding to the component; A second processing module is used to optimize the initial posture information to obtain target posture information; A third processing module is used to determine the layout result of the target space based on the target posture information; The target model is obtained through training using the category information of the sample space and the sample layout information corresponding to the sample space as training samples.

[0012] According to the spatial layout device of the present application, by learning the placement relationship between components and the relationship between furniture and room types to perform model training, the complexity of the model can be reduced so that part of the task can still be completed with a small amount of data, the convergence of the model can be improved, and on the basis of ensuring the accuracy of model training, the requirements for data and network depth are significantly reduced, the computing efficiency is improved, and it is suitable for complex and demanding layout scenarios and has universal applicability.

[0013] In a third aspect, the present application provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the spatial layout method as described in the first aspect above is implemented.

[0014] In a fourth aspect, the present application provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the spatial layout method as described in the first aspect above is implemented.

[0015] In a fifth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the spatial layout method described in the first aspect above.

[0016] The above one or more technical solutions in the embodiments of the present application have at least one of the following technical effects: By learning the placement relationship between components and the relationship between furniture and room types to train the model, the complexity of the model can be reduced so that part of the task can still be completed with a small amount of data, improving the convergence of the model. On the basis of ensuring the accuracy of model training, it significantly reduces the requirements for data and network depth, improves computing efficiency, and is suitable for complex and demanding layout scenarios with universal applicability.

[0017] Furthermore, by grouping the components output by the target model, it is possible to retain as many components with strong correlations learned in the target model and the initial pose information between components as possible in the subsequent optimization process, and to make full use of the relative relationship between the associated components. In the subsequent optimization process, there is no need to redefine fine-grained rules, and the rules learned by the target model can be directly used, which greatly reduces the number of components that need to be placed. While ensuring the accuracy and rationality of the optimization, it can also significantly improve the processing efficiency, making it suitable for complex and changeable design scenarios.

[0018] Furthermore, a heuristic algorithm is used to optimize the initial pose information. It has a faster running speed and can give a reasonable solution within a limited time. It is suitable for large-scale scenarios or complex constraints and has a higher optimization efficiency. In addition, it can flexibly adapt to a variety of different layout scenarios and is suitable for dynamically changing room structures and furniture types, with high adaptability. By designing different heuristic functions, multiple objectives can be optimized, such as aesthetics, functionality, and space utilization, to improve the layout effect.

[0019] Furthermore, when the optimization algorithm still cannot obtain the optimal solution, the components output by the target model are deleted according to the area of ​​the target space and the size of the components, so as to optimize the layout based on the remaining components, which can further improve the rationality and feasibility of the layout.

[0020] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which: Figure 1 It is a flowchart of a spatial layout method provided in an embodiment of the present application; Figure 2 is a schematic diagram of the structure of a space layout device provided in an embodiment of the present application; Figure 3 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments in the present application belong to the scope of protection of this application.

[0023] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are generally of one type, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" in the specification and claims represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0024] The spatial layout method, spatial layout device, electronic device and readable storage medium provided in the embodiments of the present application are described in detail below with reference to the accompanying drawings through specific embodiments and their application scenarios.

[0025] The spatial layout method may be applied to a terminal, and may be specifically executed by hardware or software in the terminal.

[0026] The spatial layout method provided in the embodiment of the present application may be executed by an electronic device or a functional module or functional entity in the electronic device that can implement the spatial layout method. The electronic devices mentioned in the embodiment of the present application include but are not limited to mobile phones, tablet computers, computers, cameras, and wearable devices. The spatial layout method provided in the embodiment of the present application is described below using the electronic device as an example of the execution subject.

[0027] It should be noted that the space layout method provided in the present application can be used for space layout in a virtual reality scene, and can also be used for space layout in a real scene.

[0028] like Figure 1 As shown, the spatial layout method includes: step 110, step 120 and step 130.

[0029] Step 110: input the category information of the target space into the target model, and obtain at least one component associated with the target space output by the target model, and initial pose information corresponding to the component; In this step, the target space is any space that needs to be laid out, including but not limited to virtual space or real space.

[0030] The category information is used to characterize the category of the target space, such as kitchen, living room, bedroom, office, and activity room.

[0031] Components are items placed in the target space, such as furniture, decorations or other display objects.

[0032] It is understandable that the display objects corresponding to different types of spaces and the spatial layouts corresponding to the display objects may be different; for example, the furniture included in the kitchen and the bedroom and the layouts of the furniture are different.

[0033] In some embodiments, the initial posture information may include: at least one of a relative position and a rotation relationship.

[0034] For example, the TV cabinet, wardrobe, bed and bedside table should be placed against the wall, the bedside table and bed should be placed in the same direction and the bedside table should be at the head of the bed, etc.

[0035] The target model is used to predict the component information required to generate the target space layout corresponding to the category information and the initial pose information corresponding to the component based on the category information of the target space of the required layout.

[0036] The component information includes: component category and component quantity, etc.

[0037] In some embodiments, the target model may be a deep learning model, including but not limited to: an artificial neural network (ANN), a convolutional neural network (CNN), a recurrent neural network (RNN), and a diffusion model.

[0038] The training samples include: category information of the sample space and sample layout information corresponding to the sample space.

[0039] During the actual training process, multiple sample spaces of different categories and the actual sample layouts corresponding to the sample spaces of each category can be obtained in advance, the sample components included in the actual sample layouts can be extracted, and the relative positions and rotation relationships between the sample components can be obtained. The categories and quantities of the sample components and the relative positions and rotation relationships between the sample components can be used as the sample layout information corresponding to the sample space, and multiple training samples can be constructed to train the target model.

[0040] By using deep learning algorithms to train the target model, it can automatically learn from training samples without the need to manually define rules or heuristic functions. It can flexibly adapt to various complex scenarios and component placement requirements of different styles, and has high adaptability. In addition, it can also process large-scale and high-dimensional data, which is suitable for complex three-dimensional indoor scenes and diverse component types, and improves the quality of subsequent generated layout results.

[0041] In some embodiments, training samples can also be set up based on different user styles, such as using the user's historical layout information and preferred layout data as training samples to train a target model that can predict the components of the user's preferred style and the initial pose information corresponding to the components, to achieve personalized customization and thus improve user experience.

[0042] Of course, in other embodiments, in addition to the deep learning algorithm, any other feasible training method can be used to train the target model, and this application is not limited here.

[0043] Step 120, optimizing the initial posture information to obtain the target posture information; In this step, the target posture information may include: at least one of a relative position and a rotation relationship.

[0044] The target pose information may be the same as the initial pose information, or may be different from the initial pose information.

[0045] In the actual implementation process, manual adjustment can be used to optimize the initial pose information, such as adjusting the position relationship of a component relative to the target space, or adjusting the relative positions between several components, so that the adjusted layout is more reasonable or more in line with current needs.

[0046] In some embodiments, step 120 may include: The initial pose information is optimized based on the preset constraints to obtain the target pose information.

[0047] In this embodiment, the preset constraint conditions are used to stipulate the specific placement positions of certain specific components in the target space and / or the relative positions between multiple specific components.

[0048] Preset constraints can be customized by the user, such as being set based on specific project requirements.

[0049] For example, after obtaining the component and the initial pose information corresponding to the component through the target model, it can be further determined whether the current initial pose information meets the preset constraints. If not, the initial pose information is adjusted according to the preset constraints so that the adjusted pose information meets the preset constraints, and the adjusted pose information is determined as the target pose information.

[0050] In some embodiments, step 120 may include: A heuristic algorithm is used to optimize the initial pose information and obtain the target pose information.

[0051] In this embodiment, the heuristic algorithm is an algorithm for finding a near optimal solution based on certain experience and heuristic strategies.

[0052] In furniture placement, heuristic algorithms usually design heuristic functions to gradually adjust the placement of furniture according to the current furniture placement and the constraints of the remaining space until a reasonable layout is found.

[0053] Heuristic algorithms try different solutions multiple times to get closer to the optimal solution.

[0054] For example, when the target model outputs multiple components including a TV and a sofa, the sofa should be placed opposite the TV. At the same time, the components around the door of the target space need to be adjusted so that there are no components within a certain range of the door to prevent obstruction in opening and closing the door.

[0055] In the process of optimization using heuristic algorithms, the relative positions or rotation relationships of all components are adjusted each time by guiding random iterations, and the optimal solution that meets all the above rules is gradually obtained.

[0056] For example, if the sofa is within the opening and closing range of the door, the sofa will be moved back along the door towards the sofa. If it cannot be moved back, the sofa will be transformed (such as rotating the sofa at any angle or moving it in any other direction). If the sofa partially overlaps with another piece of furniture, the sofa and the other piece of furniture will be moved back or transformed respectively. Or if the sofa needs to be placed against a wall, after each adjustment of the sofa based on the instructions of the previous step, try to place the sofa against the wall. If it fails, mutate again until all rules are met. It is approximately considered that the optimal solution is obtained, and the final posture information of the sofa and other furniture is determined as the target posture information.

[0057] Of course, in other embodiments, any other achievable optimization algorithm may be used, and this application does not limit this.

[0058] According to the spatial layout method provided in the embodiment of the present application, a heuristic algorithm is used to optimize the initial posture information, which has a faster running speed and can provide a reasonable solution within a limited time. It is suitable for large-scale scenes or complex constraints and has a higher optimization efficiency. In addition, it can flexibly adapt to a variety of different layout scenarios and is suitable for dynamically changing room structures and furniture types, with high adaptability. By designing different heuristic functions, multiple objectives such as aesthetics, functionality, and space utilization are optimized to improve the layout effect.

[0059] Step 130: Determine the layout result of the target space based on the target posture information.

[0060] In this step, the layout result is the final layout of the target space.

[0061] For any type of space, the component information corresponding to the space and the target pose information corresponding to each component can be determined based on the method of steps 110 to 120 above.

[0062] The component information includes but is not limited to component category and component quantity.

[0063] After obtaining the component information corresponding to the target space and the target pose information corresponding to each component, the components can be placed at the corresponding positions in the target space based on the target pose information, thereby completing the layout of the target space and generating a layout result that meets the actual situation and project requirements.

[0064] During the research and development process, the inventors discovered that in the related technologies, rule-based technologies lack flexibility. Since the rules are pre-defined, such systems show limitations when dealing with complex and changing scenarios, especially when dealing with users' personalized needs or non-standard scenarios. The system is difficult to adjust flexibly; when the scenario becomes more complex, adding new rules often makes the system more redundant and even causes conflicts; in addition, defining a suitable set of rules is also a time-consuming and challenging task.

[0065] In the related technology, there is also a method of obtaining spatial layout through deep learning, but this method requires a large amount of high-quality data. In the field of furniture placement, obtaining a large number of real furniture layout examples and annotating them is a time-consuming and laborious task, and the quantity and quality of open source indoor data on the Internet cannot meet the requirements, which affects the training accuracy of the model, resulting in poor convergence of the model or even non-convergence. Especially in complex furniture placement scenarios, the model may find it difficult to quickly find the optimal strategy; and in high-dimensional layout scenarios, model training may take even days, affecting layout efficiency and limited application scenarios.

[0066] In the present application, the category information of the sample space is used as the sample, and the sample layout information corresponding to the sample space is used as the sample label to train the target model. The model is trained by learning the placement relationship between components and the relationship between furniture and room types without considering the scene layout constraints, such as the shape of the indoor space, the entry and exit positions of the indoor space, what rooms are connected, etc., so that some variables can be discarded to reduce the complexity of the model so that part of the task can still be completed with a small amount of data, thereby improving the convergence of the model. While ensuring the accuracy of model training and improving the search for subsequent optimal strategies, it can also reduce the requirements for data and network depth, improve computing efficiency, and is suitable for demanding scenarios such as poor three-dimensional indoor data and small data volumes, and has universal applicability.

[0067] According to the spatial layout method provided in the embodiment of the present application, model training is performed by learning the placement relationship between components and the relationship between furniture and room types, which can reduce the complexity of the model so that part of the task can still be completed with a small amount of data, improve the convergence of the model, and significantly reduce the requirements for data and network depth on the basis of ensuring the accuracy of model training, improve computing efficiency, and be applicable to complex and demanding layout scenarios with universal applicability.

[0068] In some embodiments, step 120 may include: Grouping at least one component to obtain at least one component set; While keeping the initial pose information between the components in the component set unchanged, the initial pose information between the component sets is optimized to obtain the target pose information corresponding to each component set.

[0069] In this embodiment, a component set is a set of components that are the same or have a correlation.

[0070] Components in the same component set are associated with each other and / or belong to the same category.

[0071] For example, wardrobes of different shapes belong to the wardrobe category, and TVs and TV cabinets are components with an associative relationship.

[0072] For the same component set, the relative positions and rotation relationships between the components included therein can be determined by the initial pose information output by the target model.

[0073] It can be understood that during the training and application of the target model, the association relationship between different components and properties such as the distance can be learned synchronously, and the association relationship between components and the distance, etc. can be output synchronously while outputting the components and the initial pose information corresponding to the components.

[0074] During the optimization process, the relative positions and rotational relationships of the components included in each component set can be kept unchanged, and only the relative positions and rotational relationships between component sets are adjusted to obtain the target posture information.

[0075] In some embodiments, a heuristic algorithm may be used to optimize the initial pose information to optimize the initial pose information between each component set to obtain target pose information.

[0076] For example, through the target model prediction, it is obtained that room A includes: two wardrobes, a sofa, two beds, a TV cabinet, a TV and other components, and each component corresponds to initial posture information.

[0077] By classifying multiple components, multiple component sets such as a wardrobe group, a sofa group, a bed group and a TV group can be obtained, wherein the wardrobe group includes two wardrobes, and the TV group includes a TV cabinet and a TV.

[0078] Then, a heuristic algorithm is used based on rules and experience. For example, the TV set, wardrobe set and bed set need to be placed against the wall. If the room type is a living room, the TV set needs to be placed opposite to the sofa set; if the room type is a bedroom, the bed set and the TV set generally need to be placed opposite each other. At the same time, it is necessary to adjust the surroundings of the room door without furniture to prevent the door from getting stuck. By guiding the random iteration method, the relative position or rotation relationship of all component sets is adjusted each time, and the optimal solution that meets all rules is gradually obtained.

[0079] If the current wardrobe group is within the opening and closing range of the door, the wardrobe group will continue to retreat along the door toward the wardrobe group. If it cannot retreat, it will mutate (such as rotating the wardrobe group at any angle, or moving it in any direction). When the wardrobe group overlaps with another group of furniture, the two sets of component sets will be retreated or mutated separately. Or when you want some component sets (such as the TV group and the wardrobe group) to be able to lean against the wall, try to lean the TV group and the wardrobe group against the wall after each action. If they fail to lean against the wall, mutate until the optimal solution is obtained.

[0080] According to the spatial layout method provided in the embodiment of the present application, by grouping the components output by the target model, it is possible to retain as many components with strong correlations learned in the target model and the initial posture information between the components as possible in the subsequent optimization process, and to make full use of the relative relationship between the associated components. In the subsequent optimization process, there is no need to redefine fine-grained rules, and the rules learned by the target model can be directly used, which greatly reduces the number of components that need to be placed. While ensuring the accuracy and rationality of the optimization, it can also significantly improve the processing efficiency, which is suitable for complex and changeable design scenarios.

[0081] In some embodiments, grouping at least one component may include: receiving a first input from a user, where the first input is used to input group information; In response to the first input, at least one component is grouped based on the grouping information.

[0082] In this embodiment, the grouping information is used to characterize the association and distance relationship between components.

[0083] The distance is used to determine whether components belong to the same category. The closer the distance, the greater the possibility that the components belong to the same category.

[0084] The grouping information can be customized based on the user.

[0085] The first input is used to input grouping information.

[0086] The first input may be in at least one of the following ways: First, the first input may be a touch operation, including but not limited to a click operation, a slide operation, and a press operation.

[0087] In this implementation, receiving the first input from the user may be receiving a touch operation of the user in the display area of ​​the terminal display screen.

[0088] In order to reduce the user error rate, the effective area of ​​the first input can be limited to a specific area, such as the upper middle area of ​​the user interaction interface; or when the user interaction interface is displayed, the target control is displayed on the current interface, and the first input can be achieved by touching the target control; or the first input can be set to a continuous multiple tapping operation on the display area within a target time interval.

[0089] Secondly, the first input may be a physical key input.

[0090] In this embodiment, the terminal body is provided with corresponding physical buttons (such as a mouse or keyboard, etc.), and receiving the user's first input can be receiving the operation of the user pressing the corresponding physical button; the first input can also be a combined operation of pressing multiple physical buttons at the same time.

[0091] Third, the first input may be voice input.

[0092] In this implementation, the terminal may receive the user's voice information and perform semantic recognition on the voice information to determine the grouping information based on the semantic recognition result.

[0093] Of course, in other embodiments, the first input may also be in other forms, including but not limited to character input, etc., which can be determined according to actual needs and is not limited in this embodiment of the present application.

[0094] After the user customizes the association and distance between components, the system can respond to the first input and group all components output by the current target model based on the association and distance between components defined by the user, and divide components with association or close distance into a group to obtain a corresponding component set.

[0095] According to the spatial layout method provided in the embodiment of the present application, components are grouped by the association and distance relationship between user-defined components. Flexible grouping can be performed based on actual design requirements. The method has high flexibility and is suitable for complex and changeable design scenarios.

[0096] In some embodiments, step 120 may include: When the number of optimizations of the initial pose information exceeds the optimization threshold and the optimal solution is still not obtained, at least one component is deleted from all components based on the area of ​​the target space and / or the size of the component to obtain the target component; Optimize the initial pose information corresponding to the target component to obtain the target pose information.

[0097] In this embodiment, the optimization threshold may be based on user customization.

[0098] The target components are all components remaining after deletion.

[0099] In the actual implementation process, some components that occupy a large area and are not necessary can be deleted according to the area of ​​the room and the size of the components. For example, if the target space is a bedroom, a sofa or one of the wardrobes can be deleted.

[0100] In some embodiments, components may also be deleted by random deletion, such as randomly deleting a component when the area of ​​the room is small.

[0101] Continuing with the optimization using the heuristic algorithm as an example, if the optimal solution is still not obtained after the number of adjustments exceeds the optimization threshold, such as if the position information corresponding to the component that satisfies all rules and experience and is not within the door's opening and closing range is still not obtained, it may be that there is no optimal solution under the current conditions due to factors such as the area of ​​the room and the size of the furniture. In this case, some components can be randomly deleted according to the area of ​​the room, or some non-essential components that occupy a larger area can be deleted according to the size of the components, and all the remaining components after the deletion are used as new components (i.e., target components), and the optimization strategy of steps 110 to 130 is re-executed until the optimal solution is found.

[0102] In some embodiments, when components are grouped to obtain multiple component sets, a component set may be deleted, or one or more components in a component set may be deleted, and all components included in the remaining component sets may be used as target components.

[0103] According to the spatial layout method provided in the embodiment of the present application, when the optimization algorithm still cannot obtain the optimal solution, the components output by the target model are deleted according to the area of ​​the target space and the size of the components, so as to optimize the layout based on the remaining components, which can further improve the rationality and feasibility of the layout.

[0104] The spatial layout method provided in the embodiment of the present application may be executed by a spatial layout device. In the embodiment of the present application, the spatial layout device provided in the embodiment of the present application is described by taking the spatial layout method executed by the spatial layout device as an example.

[0105] The embodiment of the present application also provides a spatial layout device.

[0106] like Figure 2 As shown, the spatial layout device includes: a first processing module 210, a second processing module 220 and a third processing module 230.

[0107] The first processing module 210 is used to input the category information of the target space into the target model, and obtain at least one component associated with the target space output by the target model, and initial pose information corresponding to the component; The second processing module 220 is used to optimize the initial posture information to obtain the target posture information; The third processing module 230 is used to determine the layout result of the target space based on the target posture information; The target model is obtained by training using the category information of the sample space and the sample layout information corresponding to the sample space as training samples.

[0108] According to the spatial layout device provided in the embodiment of the present application, model training is performed by learning the placement relationship between components and the relationship between furniture and room types, which can reduce the complexity of the model so that part of the task can still be completed with a small amount of data, improve the convergence of the model, and significantly reduce the requirements for data and network depth on the basis of ensuring the accuracy of model training, improve computing efficiency, and be applicable to complex and demanding layout scenarios, with universal applicability.

[0109] In some embodiments, the second processing module 220 may also be used to: At least one component is grouped to obtain at least one component set; components in the same component set are associated with each other and / or belong to the same category; While keeping the initial pose information between the components in the component set unchanged, the initial pose information between the component sets is optimized to obtain the target pose information corresponding to each component set.

[0110] In some embodiments, the second processing module 220 may also be used to: receiving a first input from a user, where the first input is used to input group information; In response to the first input, at least one component is grouped based on the grouping information.

[0111] In some embodiments, the second processing module 220 may also be used to: A heuristic algorithm is used to optimize the initial pose information and obtain the target pose information.

[0112] In some embodiments, the second processing module 220 may also be used to: When the number of optimizations of the initial pose information exceeds the optimization threshold and the optimal solution is still not obtained, at least one component is deleted from all components based on the area of ​​the target space and / or the size of the component to obtain the target component; Optimize the initial pose information corresponding to the target component to obtain the target pose information.

[0113] The spatial layout device in the embodiment of the present application can be an electronic device, or a component in the electronic device, such as an integrated circuit or a chip. The electronic device can be a terminal, or it can be other devices other than a terminal. Exemplarily, the electronic device can be a mobile phone, a tablet computer, a laptop computer, a PDA, a vehicle-mounted electronic device, a mobile Internet device (Mobile Internet Device, MID), an augmented reality (augmented reality, AR) / virtual reality (virtual reality, VR) device, a robot, a wearable device, an ultra-mobile personal computer (ultra-mobile personal computer, UMPC), a netbook or a personal digital assistant (personal digital assistant, PDA), etc., and can also be a server, a network attached storage (Network Attached Storage, NAS), a personal computer (personal computer, PC), a television (television, TV), a teller machine or a self-service machine, etc., which is not specifically limited in the embodiment of the present application.

[0114] The spatial layout device in the embodiment of the present application may be a device having an operating system. The operating system may be an Android operating system, an IOS operating system, or other possible operating systems, which are not specifically limited in the embodiment of the present application.

[0115] The spatial layout device provided in the embodiment of the present application can achieve Figure 1 To avoid repetition, the various processes implemented by the method embodiment are not described here.

[0116] In some embodiments, Figure 3 As shown, an embodiment of the present application also provides an electronic device 300, including a processor 301, a memory 302, and a computer program stored in the memory 302 and executable on the processor 301. When the program is executed by the processor 301, each process of the above-mentioned spatial layout method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0117] It should be noted that the electronic devices in the embodiments of the present application include the mobile electronic devices and non-mobile electronic devices mentioned above.

[0118] An embodiment of the present application also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the various processes of the above-mentioned spatial layout method embodiment are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0119] The processor is the processor in the electronic device described in the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory ROM, a random access memory RAM, a magnetic disk or an optical disk.

[0120] An embodiment of the present application also provides a computer program product, including a computer program, which implements the above-mentioned spatial layout method when executed by a processor.

[0121] The processor is the processor in the electronic device described in the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory ROM, a random access memory RAM, a magnetic disk or an optical disk.

[0122] An embodiment of the present application further provides a chip, which includes a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned spatial layout method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0123] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0124] It should be noted that, in this article, the terms "comprise", "include" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise one..." do not exclude the presence of other identical elements in the process, method, article or device including the element. In addition, it should be pointed out that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in reverse order according to the functions involved, for example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0125] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a computer software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0126] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the claims, all of which are within the protection of the present application.

[0127] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0128] Although the embodiments of the present application have been shown and described, those skilled in the art will appreciate that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present application, and that the scope of the present application is defined by the claims and their equivalents.

Claims

1. A spatial layout method, characterized in that: include: Inputting category information of the target space into the target model, obtaining at least one component associated with the target space output by the target model, and initial pose information corresponding to the component; Optimizing the initial posture information to obtain target posture information; Based on the target posture information, determining a layout result of the target space; The target model is obtained through training using the category information of the sample space and the sample layout information corresponding to the sample space as training samples.

2. The spatial layout method according to claim 1, characterized in that: The step of optimizing the initial posture information to obtain target posture information includes: The at least one component is grouped to obtain at least one component set; the components in the same component set are associated with each other, and / or the components belong to the same category; While keeping the initial pose information between the components in the component set unchanged, the initial pose information between the component sets is optimized to obtain the target pose information corresponding to each component set.

3. The spatial layout method according to claim 2, characterized in that: The grouping of the at least one component comprises: receiving a first input from a user, where the first input is used to input group information; In response to the first input, the at least one component is grouped based on the grouping information.

4. The spatial layout method according to any one of claims 1 to 3, characterized in that: The step of optimizing the initial posture information to obtain target posture information includes: A heuristic algorithm is used to optimize the initial pose information to obtain the target pose information.

5. The spatial layout method according to any one of claims 1 to 3, characterized in that: The step of optimizing the initial posture information to obtain target posture information includes: When the number of optimization times of optimizing the initial pose information exceeds an optimization threshold and an optimal solution is still not obtained, based on the area of ​​the target space and / or the size of the component, at least one component is deleted from all the components to obtain a target component; Optimize the initial pose information corresponding to the target component to obtain the target pose information.

6. The spatial layout method according to any one of claims 1 to 3, characterized in that: The initial posture information includes at least one of a relative position and a rotation relationship.

7. A spatial layout device, characterized in that: include: A first processing module, configured to input category information of a target space into a target model, and obtain at least one component associated with the target space output by the target model, and initial pose information corresponding to the component; A second processing module is used to optimize the initial posture information to obtain target posture information; A third processing module is used to determine the layout result of the target space based on the target posture information; The target model is obtained through training using the category information of the sample space and the sample layout information corresponding to the sample space as training samples.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the spatial layout method as described in any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the spatial layout method as described in any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the spatial layout method according to any one of claims 1 to 6 is implemented.

Images