Research Article

DecorPGNet: Functional Area Division and Layout Algorithm Model in Living Rooms of Chinese Apartment-Style Family Homes

Haitong Wei*

Artificial Intelligence Human Settlement Environment Joint Laboratory, DECOR.AI and Country Garden Group, Beijing and HongHao Data Intelligence Technology Co., Ltd., and Data Intelligence Branch, Enterprise Financial Management Association of China, Beijing, China

Submission: July 25, 2024; Published: August 05, 2024

*Corresponding Author: Haitong Wei, Algorithm Scientist, Entrepreneurs, Artificial Intelligence Human Settlement Environment Joint Laboratory, DECOR.AI and Country Garden Group, Beijing and HongHao Data Intelligence Technology Co., Ltd., and Data Intelligence Branch, Enterprise Financial Management Association of China, Beijing, China, Beijing, China

How to cite this article: Haitong W. DecorPGNet: Functional Area Division and Layout Algorithm Model in Living Rooms of Chinese Apartment-Style Family Homes. Civil Eng Res J. 2024; 15(1): : 555902. DOI 10.19080/CERJ.2024.15.555902

Abstract

In the context of rapid urbanization, Chinese families have set stricter standards for functionality and comfort in the design of living rooms in apartment-style residences. This trend profoundly affects the daily experience of residents. This article has achieved innovative division and layout optimization of the living room area through smart technology. Relying on the comprehensiveness and accuracy of the DecorCGCAD floor plan dataset, combined with an innovative data structure, we has developed a standardized data preprocessing and feature extraction process, significantly improving the consistency and efficiency of data processing. We use a logistic regression algorithm, which not only enhances the transparency and credibility of the model but also significantly improves the model’s predictive accuracy and spatial insight by introducing four key features (Space Balance Ratio, Entrance Line Ratio, Kitchen Threshold Distance and Balcony Threshold Distance) through innovative feature engineering. The model’s TOP5 accuracy rate has been increased to 97.3% and the TOP1 accuracy rate has reached 86.4%, fully proving its efficiency and reliability in predicting the layout of living spaces.

Further, this article has ingeniously solved the conflict and balance between multiple objectives by applying a simulated annealing algorithm and innovative evaluation functions. The proposed Adaptive Search Region Candidacy Tactic algorithm strategy, with its excellent adaptability and flexibility to irregularly shaped spaces, greatly enhances the immediate optimization and adjustment capabilities of interior design. In terms of furniture layout strategy, we have optimized the use of space and improved the living experience by considering spatial layout, functional requirements and aesthetic principles. Moreover, through effective communication between designers and algorithmic models, we have enhanced the level of personalized design, making the design process more scientific and efficient. It is particularly worth mentioning that the panoramic rendering of the Binhu City community in Hangzhou vividly demonstrates the application of the research content in the display of interior design, providing a comfortable space suitable for family communication and entertainment. Overall, this article provides a comprehensive, scientific and practical solution for the field of interior design. It not only promotes technological advancement but also provides strong support for designers and researchers, opening up new possibilities for creating more humanized and intelligent living spaces.

Keywords: Interior Design Optimization; Logistic Regression; Simulated Annealing; Living Room Division Design; Layout Optimization; Humanized Living Spaces

Abbreviations: CGAN: Conditional Generative Adversarial Networks; CNN: Convolutional Neural Networks;

Introduction

In today’s era of continuously improving living standards, Chinese families have an increasingly pronounced demand for personalized and multifunctional living environments. The living room, as the core place for family socializing, leisure and entertainment, should not only meet basic functional needs in its design but also take into account aesthetics, comfort and the personalized needs of family members. In high-rise apartment-style residences, although the space of the living room is limited, its importance in family life is irreplaceable. Furniture, as a key element in controlling and determining the use of the living room, will have residents invest a lot of time and resources to carefully arrange according to personal preferences, needs and priorities. Therefore, the division and layout optimization of functional areas in the living room play a crucial role in enhancing the living experience and meeting diverse life needs [1].

In modern family life, family members have a variety of hobbies and living habits, which requires the living room space to be flexible to adapt to different needs. For example, young parents may need a quiet working area, while children need a spacious play area. Scientific functional area division can provide suitable activity space for each family member, meeting their personalized life needs. The space of apartment-style residences is relatively limited, so efficient use of every inch of space has become the key to design. Through reasonable functional area division, the living room can be divided into different functional areas such as the reception area, entertainment area and reading area, which not only avoids the waste of space but also improves the efficiency of space use, making the limited space play the greatest effect.

As the main place for daily interaction among family members, reasonable functional area division can promote communication and interaction among family members. For example, dividing the living room into a parent-child interaction area and an adult communication area can provide suitable activity space for family members of different ages, enhancing the emotional connection between family members. Reasonable functional area division and layout not only improve the living experience but also design a beautiful and practical living room space by considering people’s behavioral habits and visual feelings. For example, through reasonable furniture placement and color matching, a warm and comfortable living atmosphere can be created to enhance the happiness of the residents.

With the continuous progress of modern design concepts, personalization and humanization have become important considerations for living room design. Introducing modern design concepts such as open space and multifunctional furniture can make the living room design more flexible and changeable, meeting the living needs of modern families.

In addition, reasonable functional area division and layout not only affect the comfort of living but also the physical and mental health of family members. For example, dividing the living room into a quiet reading area and an active game area can provide a suitable activity environment for family members, helping to relieve stress and promote physical and mental health. With the change of family structure, such as the increase or decrease of family members, the functional needs of the living room will also change accordingly. Flexible functional area division can conveniently adjust the layout of the living room to adapt to the living needs of different stages of family life.

In the real estate market, reasonably designed and laid out houses are often more attractive. Through scientific functional area division and layout, the market value of the house can be improved, increasing its competitiveness in the market. Under the background of sustainable development, the rational use of resources and the reduction of waste have become important considerations for design. Optimizing the functional area division and layout of the living room can reduce unnecessary space occupation and resource consumption, which is in line with the concept of sustainable development.

The living room is not only a place for family members’ daily life but also a window to display family culture and aesthetics. Through the careful design of functional area division, it can reflect the cultural taste and aesthetic taste of the family and enhance the cultural and aesthetic value of the residence. The living room is also the main place for family members to receive friends and relatives. Reasonable functional area division can provide suitable space for social activities, enhancing the social experience of family members. For example, setting up a special reception area can provide a comfortable and private communication space for family members and visitors. The living scene of modern families is changeable and the living room needs to adapt to different activity needs. Through flexible functional area division, the use function of the living room can be easily converted, such as from daily leisure to festival gatherings, from family meals to work meetings, etc.

In summary, the functional area division and layout of the living room in family residences are a multi-dimensional, interdisciplinary research field. Through scientific and reasonable design, the living room can become an ideal space for family members to enjoy life and enhance feelings and also provide strong support for the overall value of the residence [2]. Therefore, this article proposes a solution called DecorPGNet. It aims to achieve intelligent division and layout optimization of the living room area through scientific methods and advanced technical means. Specifically, this article will extract the large living room area based on the given sample house type diagram data and divide it into the living room area, dining room area, hall area and corridor area to meet the basic furniture layout needs. The specific research objectives include:

i. In-depth Analysis: In-depth analysis of the use characteristics and functional needs of the living room in Chinese apartment-style family residences, extract the principles of living room functions and life planning and provide a conceptual framework for the study.

ii. Theoretical Model Construction: Based on the results of in-depth analysis, construct a theoretical model for the division of living room functional areas based on user needs. This model will consider the behavioral habits, activity needs and cultural backgrounds of different family members to ensure its scientific and practical nature.

iii. Algorithm Development: Using machine learning technology, develop a set of living room layout optimization algorithms. This algorithm will be able to automatically analyze house type diagram data, intelligently divide the living room area and provide general furniture layout design schemes. Through the application of the algorithm, achieve automation and personalization of living room design.

iv. Practical Case Verification: Select representative Chinese apartment-style residential projects as cases and apply the developed model and algorithm for actual design. Through comparative analysis and user feedback, verify the effectiveness and practicality of the developed algorithm.

The research methods of this article involve the following aspects:

i. User Demand Analysis: Study the behavioral patterns and functional needs of different family members in the living room, including but not limited to entertainment, leisure, socializing, work and learning.

ii. Space Layout Theory: Discuss the basic principles of living room space layout, such as the balance between openness and privacy, the division of dynamic and static areas and the optimization of visual and walking flow lines [3].

iii. Technical Method Research: Research how to use machine learning, geometric calculation and other technologies for data analysis, model construction and algorithm development of living room layout.

Although the content of this article has important theoretical and practical value, there are still some challenges and bottlenecks in the implementation process:

i. Difficulty in Data Acquisition: High-quality residential use data and user behavior data are the foundation of algorithm development, but these data are often difficult to obtain, especially data involving user privacy.

ii. Diversity of Personalized Needs: The living habits, cultural backgrounds and aesthetic preferences of different families vary greatly and it is difficult to design a universal algorithm that can meet the needs of all users.

iii. Complexity and Adaptability of Algorithms: The problem of living room layout optimization itself is highly complex and there are many factors to consider. Existing algorithms may have limitations when dealing with such problems.

iv. Feasibility of Practical Application: Algorithm models may be limited by practical factors such as building materials, construction techniques and budget constraints, affecting the final design effect.

In addition, in past design practices, traditional design methods often rely on the personal experience and intuition of designers. Although this method can reflect the designer’s creativity and personality to a certain extent, it inevitably lacks a certain degree of systematic and scientific nature. This method is often difficult to achieve the most optimized solution when dealing with complex spatial relationships and diverse user needs. Therefore, this article attempts to break through this limitation and introduce a more systematic and scientific design concept.

The development of modern technology, especially artificial intelligence and big data technology, has provided new tools and methods for the realization of this goal. The application of these technologies enables us to better understand user behavior patterns and preferences, thereby providing more accurate and personalized designs for the layout of the living room. For example, by analyzing the daily activities and usage habits of family members through big data, AI algorithms can help us predict and simulate the best use of living room space, achieving intelligent layout applicable to multiple scenarios [4].

Therefore, this article aims to make the living room an ideal space for family members to enjoy life and enhance feelings through scientific and reasonable design and also provides strong support for the overall value of the residence. It not only focuses on improving the scientific and systematic nature of living room design but also hopes to promote the intelligent development of living room layout through the application of modern technology, ultimately realizing the enhancement of the overall value of the residence. This can not only bring a more comfortable and convenient life experience to the residents but also open up new avenues for the development of the residential design field.

Literature Review

Theoretical Foundation

In the field of interior design, the layout and optimization of functionality have always been the focus of attention for researchers and practitioners. As society evolves and people’s lifestyles change, the demand for interior space is also constantly evolving. How to enhance the efficiency of space use and the quality of life for residents through scientific and reasonable design is an eternal topic in interior design.

Mitton, M. and Nystuen, C. explored every key component of interior architecture from the perspectives of ergonomics and daily use in 2007. Their article can serve as a guide for interior design students and early-career professionals who seek a manual for designing a livable, practical and aesthetically pleasing space. It includes hundreds of drawings and photos illustrating key concepts in interior design, as well as room-by-room coverage of applicable building codes and sustainable development standards [5]. Liu, C. Y. proposed in 2011 that the analysis of the layout of owner-occupied residences is an important part of interior design. Designers provide the best design schemes by objectively analyzing and evaluating the house types provided by the owners. The quality of the original residential layout is determined by both housing elements and environmental elements. Interior designers mainly examine the housing factors and also consider environmental factors. His main purpose is to explore the main aspects of 12 housing and environmental elements in the analysis of residential layout and to propose thoughtful suggestions [6].

Saruwono M, et al. [7] presented the results of a study on the living room of apartment-style residences in 2012. The purpose was to track the patterns and preferences of furniture arrangement in the limited space of an apartment building inhabited by Malaysian police families located in the suburbs of Kuala Lumpur. The data showed that common items in each living room were sofas, coffee tables, dining tables and chairs, small tables for televisions and shoe racks. Three activity-based areas were identified as “living,” “dining,” and “entertainment.” Despite the limited space, these families showed creativity in reflecting their personalities and lifestyles [7]. Qiugang, R. systematically evaluated the interior design of living space through the grey system theory in 2016, constructed an evaluation index system and verified its effectiveness with 10 cases [8].

In 2017, Liu, C. and Wu, J. et al. proposed a learning-based method that efficiently converts raster images to vector graphics through neural networks and integer programming, ensuring topological and geometric consistency, significantly improving accuracy and recall rate, supporting 3D model popping and architectural reshaping and has been successfully applied to 100,000 images [9]. Kán, P. and Kaufmann, H. et al. introduced a system for automatically filling and optimizing virtual indoor scenes, which uses furniture objects and is based on aesthetics, ergonomics and functional rules, optimized through a mathematical cost function and genetic algorithms. The system was further extended to automatically select furniture objects and optimize material distribution to achieve design unity and color harmony. The research verified the perceptual rationality and activatability of the interior designs generated by the system [10].

Uludağ, C. explored the importance and role of furniture in living spaces in 2018. Furniture not only provides functionality to living spaces but also meets aesthetic and ergonomic needs, giving the space a specific identity. Each piece of furniture is connected to the space and the user according to its task and location. The correct selection and placement of furniture require in-depth research and data evaluation to consider user preferences, space characteristics and actor needs to achieve the optimal living environment design [11]. In the same year, Jie Z and Xuejin C, et al. [12] proposed a system that allows non-professional users to design convertible furniture. The system generates suggested convertible furniture by inputting furniture units, solving geometric problems and ensuring functionality in both folding and extending states. Faced with complex optimization problems, an efficient algorithm based on tree selection and pre-calculated configuration space was proposed to accelerate the optimization process. Experiments showed that the system could provide design suggestions and stimulate user creativity [12].

Yang B, et al. [13] proposed an automatic indoor furniture arrangement plan in 2019, using Conditional Generative Adversarial Networks (CGAN) to divide functional areas and fill in furniture. This method optimizes the objective function through the learning process, trains a fully connected network model and achieves intelligent furniture arrangement in specific areas. Experiments showed its performance and effect are better than existing technologies [13]. Wang, W. et al. in 2020, for home decoration floor plans without text descriptions, used the YOLOv3 model for door and window detection and an improved C4.5 algorithm for room classification to achieve automatic decoration. The experimental results showed that the mAP of door and window detection reached 94.013% and the room classification accuracy rate increased to 78.71%, both better than traditional methods [14]. Lv, X. and Zhao, S. proposed a framework for automatically recognizing and reconstructing residential floor plans in 2021. By extracting room information through deep segmentation, key point detection and clustering analysis, it achieved high precision and strong generalization capability for 3D vectorization reconstruction. The system performed well in floor segmentation and sloping wall detection [15].

In 2022, Wu, W. and Feng, Y. proposed an automatic layout method for indoor space based on Convolutional Neural Networks (CNN), which defines rooms and walls through a two-stage algorithm and provides the RPLAN dataset. This method imitates the designer’s process, automatically completes the layout of indoor areas and the effect is close to that of professional designers, showing the potential of computer science and technology in the field of interior design [16].

Domestic researchers in China on the layout of living room space usually conduct research on the optimization of spatial layout, the relationship between furniture and space and the diversification of functionality. These studies provide rich theoretical support and practical guidance for interior designers in China, promoting the continuous optimization and innovation of living room space layout and functionality.

Liu, Y. et al. discussed the impact of geomancy on interior design, especially in the living room and bedroom, from the perspective of geomancy in 2010. With the widespread acceptance of geomancy again, its research and application in the field of interior design have played a positive role in enhancing the comfort and functionality of the living environment [17]. Jin, T.L. revealed the differences between South Korea and China in multi-family housing types through comparison in 2014, including differences in the layout of the entrance and living room, the connection method of LDK, the plane composition and the layout of the bathroom. The analysis of long-type, tower-type and mixed-type apartments from 2000 to 2010 found that Chinese apartments have more diversity in unit planning and shape design [18]. Zhang, Z.F. discussed the importance of living room space division in interior design in 2016, analyzed different division methods and summarized the issues to be aware of in design. With the improvement of people’s requirements for the living environment, optimizing the division of living room space is of great significance for improving the quality of interior design [19]. Zhang, Y. emphasized the functionality of space layout in interior design in 2017, believing that reasonable layout is the core of design. It pointed out that the personality and aesthetics of living places are gradually formed on the basis of meeting functional needs, indicating that function is the basis for shaping the personality and aesthetics of living space [20].

In 2023, Wang, L. conducted a study on the optimization of spatial layout and functionality in interior design, discussing the partitioning, organization and relationships of space and proposing optimization strategies such as the rational use of space, flexible layout and ergonomics. The effectiveness of these strategies was verified through case studies and challenges in design were pointed out, such as the integration of multifunctional layouts, balancing functional requirements with spatial limitations, the application of new technologies and materials and the integration of user needs [21]. In the same year, Yi, H.S. conducted research on the optimization of spatial layout and functional zoning in architectural design, emphasizing that optimizing spatial layout and functional zoning can enhance the comfort within the building, improve space utilization, enhance the flexibility and variability of the building and improve the building’s sustainability [22].

Through a comprehensive analysis of domestic and international related research, it has been found that although a large number of literature has discussed spatial layout and functional optimization, there are still many issues to be resolved. For example, how to achieve the integration of multiple functions within a limited space and how to balance aesthetic appeal and environmental sustainability while meeting functional needs. The resolution of these issues requires not only the creativity and experience of designers but also interdisciplinary cooperation and innovation.

Theoretical Gaps and Innovations

Standardization of Floor Plan Data Processing

In the field of interior design, the diversity and complexity of floor plan data have always been a challenge for the uniform processing of algorithmic models. To address this issue, there is an urgent need to develop a standardized method for data preprocessing and feature extraction. Through this method, we can ensure that the input data received by the algorithmic models is consistent and accurate. This article focuses on how to extract key features from the original floor plan materials, such as the shape, area of rooms and the location of doors and windows and transform them into a format that is easy for algorithms to process. In close cooperation with Country Garden Group, we have successfully designed and constructed an innovative standardized floor plan data structure. This achievement not only provides a solid data foundation for indoor space optimization research but also provides designers and researchers with an efficient and reliable tool.

Model Interpretability and Dynamic Adaptability

In the field of machine learning, model interpretability and dynamic adaptability are two core indicators for assessing its effectiveness. Although deep learning has made breakthroughs in multiple fields, its “black box” characteristic is often criticized for its lack of transparency. To enhance transparency and trust, this article uses the logistic regression algorithm to build a classifier model. Logistic regression provides excellent interpretability with clear decision logic and intuitive output results, thereby enhancing user trust in the model. This article emphasizes that given the continuous evolution of living space usage patterns and user needs, the model must have dynamic adaptability. The model can sensitively capture and respond to changes in living spaces, achieving self-learning and optimization to adapt to the evolution of user needs and the environment. The model proposed in this research can not only make effective layout plans based on initial design parameters but, more importantly, it can continuously self-learn and retrain the model by collecting user feedback, reflecting the continuous evolution and optimization of the model. In addition, through innovative feature engineering, the model in this article can identify and utilize key features of living spaces, thereby improving the accuracy and adaptability of predictions.

Balancing in Multi-Objective Optimization

In multi-objective optimization tasks, conflicts between different objectives are often difficult to reconcile and how to balance these objectives is a huge challenge. To achieve this, it is necessary to develop optimization algorithms that can consider multiple objectives comprehensively and study how to achieve the best design scheme. This article introduces the simulated annealing algorithm, aiming to find the optimal combination list. When constructing the evaluation function, we do not rely solely on the total score generated by the classifier but further introduce innovative strategy items and penalty items. The introduction of these new elements aims to balance the scores, ensuring that the results not only reflect performance indicators but also take into account many considerations in practical applications. Strategy items encourage the algorithm to tend to combinations that conform to specific design principles or objectives, while penalty items suppress options that may have adverse effects. Through such dynamic balance, we hope to find the best solution that not only performs well in theory but also has strong operability at the application level.

Functional Area Search in Geometric Space

In interior design practice, the irregular polygonal shape of the living room is often caused by architectural features or spatial limitations, which brings a series of challenges to the effective division of functional areas. Traditionally, area division methods are designed based on regular geometric shapes and they show obvious unsuitability when dealing with irregularity, leading to insufficient utilization of space and inability to effectively meet the functional needs of residents. In addition, existing algorithms often lack research in achieving visual balance and spatial perception and show obvious limitations when dealing with dynamic changes in space and multi-dimensional constraints (such as lighting, ventilation, acoustics, etc.). These limitations restrict the immediacy and flexibility of space optimization, making it difficult to adapt to the continuous evolution of residents’ needs. To address these challenges, this article proposes a dynamic search algorithm strategy called Adaptive Search Region Candidacy Tactic (ASRCT). It can find the optimal solution in a changing environment in real time. By continuously evaluating and adjusting the boundaries of the area, it adapts to changes in the environment or needs, achieving immediate optimization and adjustment of space. This algorithm enhances adaptability and flexibility to irregularly shaped spaces, improves the ability to optimize under multi-dimensional constraints and provides a new solution for interior design.

Personalized Needs

Existing design schemes often neglect in-depth analysis of different family structures and living habits, resulting in insufficient practicality and user satisfaction in design results. In addition, existing algorithmic models also show deficiencies in interaction and cooperation with designers. To address these issues, this article establishes a two-way communication channel between designers and algorithmic models, enabling designers to evaluate and provide feedback on the model’s output. At the same time, the model can also self-learn and optimize based on designer feedback. This interaction not only enhances the level of personalization in design but also makes the design process more scientific and efficient. In addition, the dataset in this article also carefully includes a variety of design schemes for the same type of house, which not only demonstrates the creativity and ideas of designers but also fully considers the personalized needs of users, thereby closely integrating the professional insights of designers with the actual needs of users, further enriching the diversity and depth of design solutions [23].

Conceptual Model

On the basis of reviewing domestic and foreign literature, combined with the characteristics of Chinese family style apartments, this article proposes a conceptual model for innovative division and layout optimization of the living room area, as shown in (Figure 1).

The conceptual model diagram illustrates the complete process from data collection to the final design scheme determination. It begins with obtaining foundational floor plan data from the DecorCGCAD dataset for interior design. Following this, data preprocessing is conducted to ensure the accuracy and usability of the data. Secondly, various methods are applied for data analysis and processing, including the logistic regression algorithm, which enhances the model’s interpretability. Subsequently, the ASCRT algorithm is utilized to perform real-time spatial optimization under multidimensional constraints. Then, evaluation strategies and simulated annealing functions are employed to assess and refine the design schemes. A penalty mechanism is introduced to ensure that all design proposals comply with design standards and constraints. Additionally, the placement of furniture and other elements is determined to achieve optimal space utilization. Finally, designers and algorithms are integrated to form a panoramic case display. This process encompasses multiple critical steps including data handling, model training, spatial optimization and evaluation.

Method Research

Data Sources

At the intersection of interior design and machine learning, accessing a large and diverse library of CAD drawings is a key step in developing and optimizing algorithms related to vector graphics and symbols. Such a rich data resource not only provides the necessary training and testing materials for algorithms but also greatly enhances their adaptability and accuracy in practical applications [24].

The DecorCGCAD dataset used in this article is provided by the Artificial Intelligence Human Settlement Environment Joint Laboratory (represented by lab), a comprehensive reflection of real-world residential floor plans in China. It covers 10 types of residences with 526 units, each providing multiple interior design schemes, totaling 1639 design plans. These drawings are annotated at the line granularity, ensuring the precision of the data. The CAD drawings in DWG format are not only easy to manipulate but also facilitate seamless integration with other design software, greatly enhancing the practicality and flexibility of the dataset. As shown in (Figure 2).

The diversity and comprehensiveness of the DecorCGCAD dataset make it an ideal platform for studying the automation and intelligence of interior design. This dataset not only showcases the creativity and ideas of designers but also fully considers the personalized needs of users, promoting technological progress and innovation in the field of interior design. With the support of these high-quality data, we can develop smarter and more efficient algorithms, bringing more possibilities and value to interior design.

In August 2019, DECOR.AI signed a milestone strategic cooperation agreement with Country Garden Group, jointly establishing the lab. This move marks a deep cooperation between the two parties in technological innovation and industrial upgrading, aiming to promote the application and development of artificial intelligence technology in the field of human settlements. Mr. Wei, the former Chief Algorithm Scientist of the lab, said that the core task of the lab is to build a bridge connecting the needs of real estate developers for one-stop solutions with the supply capacity of high-quality home products from Foshan, China. Through this platform, we can not only meet the developers’ demand for efficient and personalized residential solutions but also provide home companies with an excellent opportunity to showcase their products and achieve intelligent manufacturing transformation.

The establishment of the lab not only strengthens the synergy between industries, promoting technological innovation and industrial upgrading, but also has far-reaching social significance in promoting the intelligent and green development of the entire home industry. It will help improve the quality and efficiency of home products, reduce production costs and enhance the market competitiveness of enterprises. At the same time, by using intelligent technology, it can better meet consumers’ needs for a healthy, comfortable and convenient living environment, improving people’s quality of life.

In addition, the lab will also be committed to researching and developing cutting-edge artificial intelligence technologies, including but not limited to machine learning, big data analysis, intelligent control, etc., in order to achieve breakthroughs in the planning, design, construction and management of the living environment and create more value for society. Through this platform, we look forward to inspiring more innovative thinking, training industry talents, promoting the deep integration of technology and humanities and contributing to the construction of a smarter and more sustainable future living environment.

The accuracy and comprehensiveness of the dataset are due to the Country Garden Design Institute’s more than 6,000 professional employees distributed across 32 comprehensive institutes nationwide, forming a widely covered design network. This scale is extremely rare in the industry, ensuring that design tasks can be completed with high quality and efficiency. The design institute brings together professionals in planning, architecture, decoration, landscape, municipal engineering and other fields, achieving rapid internal professional docking and collaboration.

Rely on 26 years of industry experience, Country Garden Design Institute has developed more than 700 domestic towns and laid out in overseas markets such as Malaysia and Australia. The design team has formulated product family type standards that match the regional characteristics of different countries and regions. Through long-term accumulation, the design institute has established a group family type library containing more than 90 items in 53 categories, as well as a regional family type library covering more than 460 items in more than 60 regions. These products have been polished repeatedly to meet the needs of different climates and regional habits.

The refined control of the Country Garden Design Institute is the key to its fast drawing. All planning work of the design institute is advanced, achieving the pre-positioning of work. Before obtaining the land, the various business departments of the design institute have already started to follow up and plan in advance synchronously, ensuring that the planning map can be produced on the day of obtaining the land. In addition, the design institute updates and upgrades the family type library every year, developing an average of 4-5 new family types per month to adapt to changes in market and consumer demand.

Country Garden Group, as one of China’s leading real estate development companies, has always been committed to improving its service quality and efficiency through technological innovation. In 2019, the group demonstrated its latest achievements in the field of Building Information Modeling (BIM) - more than 500 high-precision CAD family maps. These drawings not only show the various types of family structures in China but also reflect the deep strength of Country Garden in architectural design and planning.

The source data used by DecorCGCAD is this batch of CAD family maps. It covers a variety of building types, from single-family homes to high-rise apartments, from traditional courtyards to modern villas. It includes ten types of family types: one-bedroom, two-bedroom, three-bedroom, four-bedroom, large flat, duplex, staggered, townhouse, detached and special-shaped. Each drawing is carefully designed to ensure high precision and detail performance. The drawings not only contain the basic structure of the house, such as walls, doors and windows, stairs, etc., but also clearly mark the size and material of various architectural elements, providing an accurate reference for subsequent construction and decoration.

In order to further optimize the use of these CAD family maps, we have converted these drawings into structured data files. This conversion process involves extracting the graphics and text information in the drawings and converting them into a format that can be recognized and processed by computer programs. In this way, designers and engineers can more easily search, analyze and modify data, thereby improving work efficiency and design quality.

The following is the hierarchical information of the structured data, including all fields and classifications:

1) Top-level structure

a. HouseId: The unique identifier for a residence.

b. HouseType: Residential type.

c. HouseLayout: An array containing floor and room layouts.

d. SplitSpaces: The division of different functional areas within a residential building.

e. FuncAreas: Detailed information about functional areas.

f. DoorAndWindows: Detailed information about doors and windows.

g. Components: Fixed components in residential buildings.

h. PointElements: Point elements in residential buildings.

2) House Layout (HouseLayout)

a. StoreyId: Floor identification.

b. StoreyHeight: Floor height.

c. RoomInfos: An array of room information.

d. RoomId: The unique identifier of a room.

e. RoomName: Room name.

f. RoomType: Room type.

g. JointPoints: an array of connection point information.

h. StartPoint: The starting coordinate of the connection point.

i. EndPoint: The end coordinate of the connection point.

j. Id: Unique identifier of the connection point.

k. Type: Connection point type.

l. Subtype: Subtype.

m. IsMainDoor: Is it the main door.

n. NextRoomId: The ID of the next room.

o. NextRoomName: The name of the next room.

p. NextRoomType: The type of the next room.

q. IsOpenRight: Is the right side open.

r. IsAlongLine: Is it along a straight line.

s. RadiusMidPoint: Coordinates of the midpoint of the arc.

t. RadiusLength: The length of a circular arc.

u. ObjectHeight: Object height.

v. ObjectToFloor: The height from the object to the floor.

3) Split Spaces (splitSpaces)

a. SpaceID: The unique identifier of a space.

b. SpaceName: The name of the space.

c. SpaceType: Space type.

d. ParentRoomId: The ID of the parent room.

e. JointPoints: an array of connection point information (as above).

4) Functional Areas (funcAreas)

a. AreaID: Unique identifier of the functional area.

b. AreaName: The name of the functional area.

c. AreaType: Functional area type.

d. JointPoints: an array of connection point information (as above).

5) Door and Window Information (doorAndWindows)

a. ObjectId: Unique identifier for doors and windows.

b. ObjectName: Door and window name.

c. HighToFloor: Height to floor.

d. Size: Dimensions (width, height, depth).

e. Position: Position coordinates.

f. Type: Type.

g. Rotate: Rotation angle.

h. BayWindowPoint: bay window point.

i. RadiusMidPoint: Coordinates of the midpoint of the arc.

j. RadiusLength: The length of a circular arc.

6) Components Information (components)

a. Id: Unique identifier of the component.

b. Name: Component name.

c. Type: Type.

d. IsCylinder: Is it cylindrical.

e. Vertexes: vertex coordinate array.

f. Cylinder Center: Coordinates of the center point of a cylinder.

g. Cylinder Height: The height of a cylinder.

h. Cylinder Radius: The radius of a cylinder.

i. Rotation: Rotation information.

7) Point Elements (Point Elements)

a. Id: Unique identifier of the positional element.

b. Name: Point name.

c. Type: Type.

d. IsRound: Is it a circle.

e. Vertexes: vertex coordinate array (empty if circular).

f. RoundCenter: Coordinates of the center point of a circle.

g. RoundRadius: The radius of a circle.

h. Rotation: Rotation information.

An example of structured data is shown in Figure 3

The standardization and formatting of data provide designers with a clear framework, enabling them to easily count and compare the structural characteristics of different family types. This comparison not only helps to optimize the design scheme but also promotes the innovation and diversity of design ideas. Designers can adjust and improve the design based on these analysis results to better meet user needs. At the same time, the introduction of structured data files also facilitates the application of automated tools. For example, automated tools can use these data files to automatically generate construction drawings, budget reports, etc., which not only reduces manual operation errors but also significantly reduces time costs. This automated processing not only improves work efficiency but also ensures the accuracy and consistency of construction drawings and budget reports. More importantly, the use of structured data files allows staff from different departments to work based on the same data source. This unified data source ensures the consistency and accuracy of information, avoiding misunderstandings and errors caused by inconsistent information. This is crucial for cross-departmental collaboration and project management, helping to improve the execution efficiency and quality of the entire project. Finally, the dataset in this article also includes a variety of design schemes for the same family type, which not only shows the creativity and ideas of designers but also fully considers the personalized needs of users. In this way, we not only enhance the level of personalized design but also make the design process more scientific and efficient. This structured and diverse dataset provides a rich resource library for designers and users, further promoting innovation and development in the field of interior design.

Feature Engineering

The living room is not only a place for daily family activities but also an important space for receiving friends and social interaction. Its design and functional zoning directly affect the comfort of living and the convenience of family life. Therefore, the design of the living room should not only meet the basic living needs but also consider the emotional communication and social needs of family members. This requires designers to fully consider the multifunctionality of the living room in the design, so that it can meet the needs of daily living and adapt to different social occasions.

In modern home design, the design and functional zoning of the living room is a crucial link. A reasonable design can not only improve the quality of life of the residents but also reflect their aesthetics and taste. Designers need to find a balance between aesthetics, functionality and the living habits of the residents to create a space that is both beautiful and practical. This includes but is not limited to the layout of the living room, the selection of furniture, the color matching, the design of lighting, etc. This requires designers to have professional design knowledge and skills, as well as keen observation and rich imagination to better understand and meet the needs of the residents.

Reasonable functional zoning can not only improve the practicality of the space but also enhance the living experience of the residents. Through scientific and reasonable design, the living room can be divided into different functional areas, such as a rest area, entertainment area, dining area, etc. This can not only improve the utilization rate of the space but also allow residents to easily switch between different activities and enjoy a more comfortable and convenient life.

The design of the living room should not only meet basic functional needs but also reflect the personality and taste of the residents. Designers can choose design styles and elements suitable for their personality through in-depth understanding of the residents. For example, residents who like modern minimalist styles can choose simple lines and colors; while those who like traditional styles can choose classical furniture and decorations. Such a design can not only improve the satisfaction of the residents but also allow them to feel their unique personality and taste in daily life.

To achieve a living room design that is both aesthetically pleasing and practical, designers need to consider a variety of factors. This includes but is not limited to the area of the area, the boundaries of the area and the relationship with the entrance door, kitchen door and balcony door. These factors not only affect the visual aesthetics of the space but also involve the daily activities and living habits of the residents. Designers need to find the best design scheme through careful observation and analysis to ensure that every corner of the living room can exert its maximum functionality and aesthetic value.

In-depth research on the characteristics of the functional zoning of the living room and quantifying them into operable mathematical models has important theoretical and practical significance for optimizing the functional zoning of the living room. Through mathematical models, designers can more scientifically analyze and optimize the spatial layout to ensure that every design decision is supported by data and theory. This not only improves the scientific and rational nature of the design but also provides possibilities for the automation and intelligence of interior design.

In the field of interior design, the cooperation between professional designers and algorithm engineers often produces surprising innovative results. The integration of feature engineering in machine learning with interior design is an interesting and creative idea. Although these two fields may seem unrelated on the surface, they can actually complement each other [25]. In this article, we will provide corresponding mathematical calculation formulas to help designers make more scientific and reasonable decisions in the design process. These formulas can help designers calculate and optimize the spatial layout more accurately to ensure that the design meets aesthetic requirements and functional needs. By in-depth research and quantification of the characteristics of the functional zoning of the living room, we hope to provide new perspectives and methods for the field of interior design and promote innovation and development in interior design. This interdisciplinary cooperation can not only promote the innovation of design methods but also greatly improve the efficiency and quality of the design. The following are the dimensions of the available model features that have been jointly studied and collided by professional designers and algorithm engineers. These dimensions have important application value in interior design.

Space Balance Ratio

In the functional zoning of the living room, the space balance ratio is a key quantifiable indicator. First, this ratio directly affects the spatial distribution of various functional areas in the living room. For example, a spacious rest area may need to occupy a larger area to allow family members to rest and relax comfortably. A small reading area can occupy a smaller area to meet the needs of reading and learning. Designers need to reasonably allocate the area of each functional area according to the living habits and preferences of the residents to ensure that each area is neither too crowded nor too empty. Second, by calculating the space balance ratio, designers can better balance the spatial needs of each functional area. This balance is not only related to the visual effect of the space but also involves the user experience of the residents. A coordinated and balanced spatial layout can avoid affecting the overall visual effect and user experience due to the functional area being too large or too small. In addition, this ratio also helps to evaluate the fluidity and openness of the space. An open spatial layout can enhance the fluidity of the space, allowing residents to move and act freely in the living room. Designers can create an open and orderly space by adjusting the area and position of the functional areas. Such a spatial layout can not only improve the quality of life of the residents but also enhance the practicality and comfort of the space. Furthermore, a reasonable space balance ratio can also enhance the practicality and comfort of the space. Through scientific and reasonable spatial planning, designers can ensure that each functional area can exert its maximum functionality and aesthetic value. Finally, a reasonable space balance ratio can allow residents to enjoy a more free and flexible activity space in the living room. Designers need to understand the daily activities and needs of the residents through careful observation and analysis, so as to create a living room space that is both beautiful and practical. Such a space can not only meet the basic needs of the residents but also allow them to feel the convenience and comfort brought by the design in daily life. Quantitative formula as follow:

Where S_regionrepresents the area of the region and S_{living room}represents the total area of the living room.

Entrance Line Ratio

The entrance line ratio is an important design consideration in the functional zoning of the living room. This ratio not only affects the visual aesthetics of the space but also involves the daily activities and living habits of the residents. Generally, designers prefer to avoid the functional area boundary facing the entrance door directly, to avoid visual abruptness and spatial inconvenience. By calculating this feature, designers can better plan the layout of the functional areas to ensure the smoothness and privacy of the space. In addition, this ratio also helps to evaluate the openness and permeability of the space, to avoid affecting the activities and mood of the residents due to improper functional area layout. A reasonable entrance line ratio can make the living room space appear more harmonious, enhancing the residents’ satisfaction and sense of belonging to the space. Quantitative formula as follow:

Where L_{entrace door}represents the length of the entrance door, L_regionrepresents the length of the regional boundary and  represents the intersection in geometry.

Kitchen Threshold Distance

The kitchen threshold distance is another key factor affecting the functional zoning of the living room, playing a crucial role in interior design. In the design process, designers can scientifically plan the layout of the functional areas by calculating the distance between the regional boundary and the kitchen door. This planning can not only ensure the coordination of the space but also enhance the comfort experience of the residents. For example, a reasonable distance can prevent the noise and smell during cooking from directly affecting the living room area, while also facilitating free movement and interaction between family members in the kitchen and living room. In addition, this distance also helps to evaluate the fluidity and openness of the space. An open spatial layout can enhance the fluidity of the space, allowing residents to move and act freely in the living room. Designers can create an open and orderly space by adjusting the distance between the regional boundary and the kitchen door. Such a spatial layout can not only improve the quality of life of the residents but also enhance the practicality and comfort of the space. Quantitative formula as follow:

Where x_region and y_region represent the x and y coordinates of the regional boundary respectively and x_{kitchen door}y_{kitchen door}represent the x and y coordinates of the kitchen door respectively.

Balcony Threshold Distance

The distance between the boundary of a living area and the balcony door plays an essential role in the functional zoning of the living room. During the design process, designers can scientifically plan the layout of the functional areas by calculating the distance between the area boundary and the balcony door. This planning ensures not only the coordination of the space but also enhances the comfort experience for the residents. For instance, a reasonable distance can prevent the living room area from being directly exposed to intense sunlight while ensuring sufficient natural light enters, creating a bright and cozy living environment. Moreover, this distance also helps assess the openness and permeability of the space. An open spatial layout can enhance the fluidity of the space, allowing residents to move and act freely within the living room. Designers can adjust the distance between the area boundary and the balcony door to create a space that is both open and orderly. Such a spatial layout can improve the quality of life for residents and enhance the practicality and comfort of the space. A reasonable distance setting can make the living room space appear brighter and more ventilated, enhancing the residents’ sense of comfort and pleasure. For example, a distance that is too close might make residents feel cramped, affecting their freedom of movement within the living room; whereas a distance that is too far might result in a lack of sufficient natural light and fresh air in the living room area. Therefore, designers need to ensure the practicality of the space while also considering the psychological feelings and emotional needs of the residents. By deeply studying and quantifying the distance between the area boundary and the balcony door, designers can better understand and meet the needs of the residents, creating a living room space that is both aesthetically pleasing and practical. In summary, the distance between the area boundary and the balcony door is a design detail that should not be overlooked. It is related not only to the lighting and ventilation of the space but also to the daily activities and psychological feelings of the residents. Quantitative formula as follow:

Where x_region and y_region represent the X and Y coordinates of the regional boundary respectively and x_{balcony door} y_{balcony door} represent the X and Y coordinates of the balcony door respectively.

Classifier Model

Logistic Regression Principle

Logistic regression, also translated as “logarithmic probability regression”, is a common algorithm for binary classification tasks [25]. It is a supervised statistical learning method and is mainly used to classify samples. Unlike linear regression, the linear regression model is used to find a function, f(x)=+b xω, such that f(x) is as close as possible to the true value. Logistic regression, on the other hand, produces and outputs a marker, y (which has a Boolean value of 0, 1), on a binary classification task. The logistic regression is generalized linear regression, which means that for a regression task, if the output of the mark changes on an exponential scale, we can use as the target for the model to predict the output approximation. Similarly, logistic regression is to find a function, g(z), to convert the real value of the predicted output of the linear regression into a Boolean value. Usually we use the sigmoid function, which takes the following form:

The function is an S-shaped function. When the independent variable, z, approaches positive infinity, the dependent variable g(z) approaches 1 and when z approaches negative infinity, g(z) approaches 0. Any real number maps to the (0,1) interval, making it useful to transform an arbitrary-valued function into a function more suitable for binary classification, which is a good fit for our classification probability model. Because of this property, the sigmoid function is also regarded as a method of normalization. Similar to normalization, it is a “scaling” function in data preprocessing, which can compress data into [0,1]. The difference is that after normalization, 0 and 1 can be taken (the maximum value is 1 after normalization and the minimum value is 0 after normalization), but the sigmoid function is only infinitely close to 0 and 1. Also, it has a nice derivative property, as follows:

After understanding the model of binary classification and regression, we have to look at the loss function of the model. Our goal is to minimize the loss function to obtain the corresponding model coefficient, θ.

Since linear regression is continuous, the loss function can be defined using the sum of the squares of the model errors. However, logistic regression is not continuous and the experience of defining the loss function in natural linear regression cannot be used. However, we can use the maximum likelihood method to derive our loss function.

According to the definition of binary logistic regression, it is assumed that our sample output is 0 or 1. Then, we have the following:

These two formulas are then combined into the following single formula:

The value of y can only be 0 or 1.

With the probability distribution function expression of y, we can use the likelihood function maximization to solve for the model coefficient θ that is needed.

Here, we used the maximum likelihood estimation method. This method was first proposed by the German mathematician C. F. Gauss, in 1821, but this method is usually attributed to the British statistician R. A. Fisher. Maximum likelihood estimation is just an application of probability theory in statistics and it is a method of parameter estimation. This means that a random sample is known to satisfy a certain probability distribution, but the specific parameters are not clear. Parameter estimation involves conducting several experiments, observing the results and using the results to deduce the approximate value of the parameters. Maximum likelihood estimation establishes that a certain parameter can make the probability of a sample appear to be the largest. Of course, we will not choose other samples with small probabilities, so we simply use this parameter as the real value of the estimate.

In order to facilitate the solution, here we maximize the log-likelihood function and the inverse of the log-likelihood function is our loss function, () J θ . The algebraic expression of the likelihood function is as follows:

where E is an all 1 vector.

To minimize the loss function in binary logistic regression, there are many methods and the most common are the gradient descent method, coordinate axis descent method, quasi-Newton method, etc., with which we derive the formula for each iteration of θ using gradient descent. Because of the tedious derivation of the algebraic method, the matrix method is used to optimize the loss function. Here, the process of deriving the gradient in binary logistic regression by the matrix method is provided.

For the above loss function expressed by the matrix method, we can use the derivation of the for the θ vector to obtain the following:

In this step, we used the chain rule of vector derivation and the matrix form of the following three basic derivation formulas (g(z) is a sigmoid function):

where α is the step size of the gradient descent method.

Training

Logistic Regression and Support Vector Machine (SVM) are two commonly used classification algorithms. Considering that Logistic Regression has fewer model parameters, higher computational efficiency, easy to calculate probabilities, easy to interpret model results and less strict requirements on the distribution of data compared to SVM [27]. This article employs the logistic regression algorithm to train the classifier for optimizing the division of different functional areas. Below is a detailed explanation of the parameters we have selected, which collectively contribute to our model, ensuring its accuracy and efficiency when dealing with complex interior design issues.

Specifically, for the Chinese apartment-style family home living room functional area division problem, we have utilized the logistic regression algorithm from the scikit-learn library (1.3.2 version), which is widely acclaimed in the field of machine learning for its stability and efficiency. Table 1 is a detailed explanation of the parameters we have selected, which collectively contribute to our model, ensuring its accuracy and efficiency when dealing with complex interior design issues.

In terms of model performance, not only have we achieved significant results in quantitative metrics, but we have also demonstrated its superiority across multiple dimensions. Specifically, the model’s TOP5 accuracy rate has reached 97.3%, which means that when recommending living space layouts that users might be interested in, the model can capture user preferences with an extremely high probability, ensuring that at least five recommended layouts meet user expectations. The TOP1 accuracy rate of 86.4% further illustrates the model’s high reliability in accurately predicting the user’s most preferred living space layout.

Such a high accuracy rate of the model is inseparable from the quality and diversity of the dataset, providing the model with rich learning and generalization capabilities. The model adopts high-quality feature engineering methods during the design, extracting features that are crucial for predicting living space layouts. During the model construction process, we continuously optimized the algorithm structure, adopted regularization strategies, effectively avoiding overfitting and improved the model’s generalization ability. The model integrated a user feedback mechanism during the iteration process, adjusting parameters in real-time to better adapt to the actual needs and preferences of users. Combining these factors, the model not only showed excellent performance in experiments but also has high practical value in practical applications.

General Mathematical Definition

The weight coefficient matrix W is an M×N matrix, where M is the number of classes and N is the number of features. Each row w_jof the matrix represents the weight vector for class j, which is dotted with the feature vector x, reflecting the contribution of the features to the prediction of class j. The sign and magnitude of the weight coefficients directly affect the influence of the features on the class probabilities.

The intercept vector b is an M-dimensional vector, where each element b_j is the intercept term for class j. The intercept term allows the model to have a non-zero baseline prediction for the class probabilities when the feature values are zero.

Based on the trained logistic regression model, this article uses the `predict_proba` method from the scikit-learn library to predict the feature vector x of new samples, obtaining the probabilities for each class. Here is the mathematical description of this process:

i. Linear Model Calculation: For each class j, the model first calculates the dot product of the feature vector x and the weight vector w_j , then adds the intercept b_j , to obtain the log odds z_j :

Region of Interest

The Adaptive Search Region Candidacy Tactic (ASRCT) is utilized for dynamically searching candidate areas with the aim of finding optimal solutions in real-time within changing environments. Central to this strategy is its “adaptive” capability, allowing it to respond flexibly to the dynamic demands of interior design and space management. Particularly crucial in the division of living room areas, this strategy involves continuous optimization and adjustment of spatial layouts.

In practice, ASCRT undergoes a series of complex calculations and evaluation processes to dynamically adjust the boundaries of living room areas. The algorithm can identify and analyze residents’ behavior patterns, furniture layout, lighting conditions and functional space requirements, thereby achieving real-time optimization of area division. For instance, if residents prefer social activities in the living room, the algorithm may correspondingly expand the resting area while adjusting other areas to maintain overall coordination and balance.

ASRCT also takes into account the physical limitations of a space, such as the positions of walls, doors and windows and other fixed structures. Through precise geometric calculations, the algorithm identifies potential area boundaries and assesses the feasibility of these boundaries in practical application. This approach not only improves the efficiency of space utilization but also enhances the humanization and adaptability of the design.

Geometric Calculation Steps:

vii. Evaluation and Selection: Evaluate the validity of all constructed candidate rectangles R_kl according to design requirements and spatial functions. Select valid rectangles R_kl that meet the conditions as partition boundaries.

viii. Dynamic Updating: Repeat the above steps with changes in the environment or requirements to update the partition scheme in real-time.

The Figure 4 is the pseudocode for ASCRT, providing a high-level description of how to generate candidate solutions for area boundary boxes from the vertices of the room contour.

The geometric computation of ASCRT is not only a technical method but also an innovative design concept. It emphasizes the dynamism, adaptability and user-centeredness of design, bringing new ideas and approaches to the field of interior design. With the continuous advancement of technology, this method is expected to combine with more advanced technologies to further promote innovation and development in interior design.

Multi-Objective Optimization

Multi-objective optimization refers to the process of considering multiple objective functions in a single optimization process. These objective functions are typically in conflict with each other and the goal is to find a balance between them. For instance, in engineering design, one may need to consider cost, efficiency and reliability simultaneously; in resource allocation, one might need to consider fairness and efficiency at the same time. The aim of multi-objective optimization is to find a set of solutions that optimize each objective function as much as possible or find an acceptable balance among them.

Simulated Annealing is a probabilistic heuristic search algorithm that originates from the annealing process in solid-state physics. In solid-state physics, annealing involves heating a material to a high temperature and then cooling it slowly to reduce defects and stress within the material, thereby achieving a better crystal structure [28]. The Simulated Annealing algorithm borrows from this process, using the simulation of temperature reduction to find the global optimum solution to a problem.

The core idea of the Simulated Annealing algorithm is to use a temperature parameter T to control the acceptance criterion of the search process. At high temperatures, the algorithm is more likely to accept worse solutions to escape from local optima; at low temperatures, the algorithm tends to accept better solutions to stabilize near the global optimum.

Algorithm Procedure:

i. Initialization: Set the initial temperature T0, initial solution S0, cooling factor α , termination temperature T_minand weight coefficient λ

Figure 5 is a high-level pseudocode description of the Simulated Annealing algorithm, which detailed shows the optimization process of living room partition combination evaluation.

Figure 6 meticulously depicts the optimal partitioning scheme that we have derived through careful analysis and rigorous selection. In this captivating visual feast, the red rectangle immediately catches the eye, representing the warm and cozy living room area, providing a comfortable space for interaction between visitors and family members.

Furniture Layout

After the functional zoning of the living room, hallway, dining room and corridors is completed, determining the placement lines, positions and directions of the main furniture in each area is a task that requires careful consideration. This article starts with the determination of furniture placement lines and describes in detail the placement strategies for sofas, TVs, dining tables and shoe cabinets in the living room.

Furniture placement lines refer to the planning of the relative positions and boundary lines of furniture according to functional needs and visual aesthetics, in order to achieve rational use of space and aesthetic layout. It mainly includes four placement lines:

i. Sofa Placement Line: First, identify the focal element of the living room, such as a fireplace or large artwork and place the sofa facing or around this focal point to enhance the cohesion of the space. Second, the sofa should facilitate interaction with visitors, considering the entrance position of the living room, making the sofa the first stop for welcoming guests. In addition, ensure that the sofa position provides a broad view while avoiding direct exposure to strong light sources to reduce glare.

ii. TV Placement Line: First, the TV should be placed opposite the sofa placement line, ensuring that the distance from the sofa to the TV is appropriate to avoid visual fatigue when watching. Second, the TV wall should avoid direct light to reduce screen glare and improve viewing comfort. In addition, the center point of the TV should align with the line of sight height of the sofa placement line, usually about 1-1.2 meters above the ground.

iii. Dining Table Placement Line: First, the placement line of the dining table should consider its convenient connection with the kitchen for easy serving and cleaning. Second, if there is no obvious separation between the dining area and the living room, the placement line of the dining table should be as far away from the main passage as possible to avoid obstructing the smoothness of the corridor. In addition, the shape and size of the dining table should match the geometric shape of the dining space to ensure spatial coordination.

iv. Shoe Cabinet Placement Line: First, the shoe cabinet should be placed in the hallway area, near the entrance, for convenient shoe changing for visitors and family members. Second, the placement line of the shoe cabinet should be parallel to the wall to reduce the occupation of the corridor space. In addition, the design of the shoe cabinet should provide enough storage space while maintaining a consistent appearance with the decorative style of the hallway.

The position and direction of furniture placement refer to determining the specific placement point and orientation of furniture indoors according to spatial layout, functional requirements and aesthetic principles, in order to optimize space use and enhance the living experience. The sofa should be placed in the central area of the living room to form a space for rest and social interaction. Ensure that the placement of the sofa does not obstruct the passage while leaving enough space for activities. The TV cabinet or wall should be perpendicular to the sofa placement, with the TV facing the center of the sofa, maintaining a direct distance between the sofa and the TV, generally recommended to be five times the diagonal length of the TV screen. The dining table should be placed in the center of the dining area, considering the comfort during dining and the flow of space, with enough space around the dining table for people to move. The shoe cabinet should be placed against the wall near the entrance of the hallway to avoid occupying corridor space and be easy to use.

In the process of carefully planning the placement of furniture, it is crucial to consider multiple key factors to create a living space that is both practical and aesthetically pleasing. First, ensure that the design of spatial flow lines is reasonable to avoid any furniture placement obstructing the main action paths, thus maintaining the fluidity and usability of the space. Second, pursue visual balance by reasonably distributing furniture to avoid the space appearing too crowded or empty, creating a harmonious and attractive visual effect. In addition, according to the specific functional requirements of each area, carefully select the type, size and placement method of furniture to ensure that it is both practical and aesthetically pleasing. Finally, consider environmental adaptability, fully integrating natural light, indoor lighting and ventilation conditions, so that the placement of furniture not only complements the indoor environment but also effectively enhances the quality of life for residents.

By comprehensively considering the above factors, the placement lines, positions and directions of the main furniture in each functional area of the living room can be effectively determined, creating a living space that is both practical and aesthetically pleasing [29].

Visual Rendering

Combining visual rendering with a knowledge graph and using panoramic images for display can bring innovative experiences to fields such as interior design, architectural visualization and even game design [30]. Static panoramic images are a type of interior design display that is fixed in perspective and does not change with the observer’s viewpoint. They are typically presented in the form of photographs or digital images. Compared to dynamic panoramic images, static panoramic images do not include any changes in viewpoint or interactivity. They provide a fixed angle to showcase a specific perspective or view of the interior design. Static panoramic images are used to display the visual effects of interior design, often including elements such as spatial layout, color, texture, lighting and decoration. Because they are static, viewers can easily view and understand the design at any time. This type of image is suitable for quickly conveying design intent and style.

In Hangzhou, China, Country Garden Group is renowned for its outstanding real estate development projects. Among them, the much-anticipated Binhu City housing estate, with its innovative floor plan designs and excellent location, has become a highlight in the local real estate market. To more intuitively display the characteristics and advantages of these floor plans, this article has specially produced static panoramic images to allow potential buyers and investors to understand the spatial layout and design details of each floor plan at a glance. As shown in Figure 7.

This image presents a modern interior design in a panoramic perspective, allowing us to fully experience the layout, color matching and detailed decoration of the space. The panoramic image not only shows the interrelationship of various functional areas but also makes us feel as if we are in this space, feeling its unique atmosphere and design concept.

From the panoramic image, it can be seen that the overall color tone of the room is mainly white and gray. The combination of these two colors creates a simple yet elegant atmosphere. White walls and ceilings make the space appear more spacious and bright, while gray serves as an accent, adding a sense of hierarchy and depth to the space. In addition, the room also cleverly incorporates wooden elements, such as the wooden dining table, bookshelf and wooden frames of some furniture. These wooden elements not only add a touch of natural atmosphere to the space but also bring a warm touch, making the entire environment more livable.

The panoramic image clearly shows the various functional areas within the room. The dining area is centered around a round dining table and four chairs, creating a warm dining environment. The living room area is the core area for family entertainment and leisure, with a combination of sofas, armchairs and a round coffee table providing a comfortable space for family and friends to communicate. From the panoramic image, it can be seen that the placement of the sofa and chairs considers both the utilization rate of the space and the visual balance. In addition, the living room area also provides sufficient lighting for the space through lighting fixtures such as chandeliers, making the entire area brighter and warmer.

The panoramic image also shows many detailed decorations in the room, such as books and decorations on the bookshelf, the pattern and color of the carpet and wall hangings. These detailed decorations not only enhance the taste and style of the space but also show the owner’s attitude towards life and aesthetic taste. For example, the books on the bookshelf may reflect the owner’s reading habits and interests; the pattern and color of the carpet are coordinated with the overall color tone, adding a touch of warmth to the space; the wall hangings may express the owner’s artistic pursuit and emotional sustenance in the form of abstract or figurative expressions.

In summary, this interior design panoramic image, through its unique perspective and meticulous depiction, allows us to fully and deeply understand this modern-style interior space. It not only shows the aesthetics and functionality of the space but also conveys a lifestyle and aesthetic taste.

Conclusion

This article has successfully achieved intelligent division and layout optimization of the living room area through the comprehensive application of scientific methods and advanced technical means. Through in-depth analysis and research, this article has proposed a series of innovative solutions that not only improve the scientific and practical nature of interior design but also provide designers and researchers with powerful tools and methods.

Firstly, in response to the diversity and complexity of floor plan data in the field of interior design, this article has developed a set of standardized data preprocessing and feature extraction methods. Through close cooperation with Country Garden Group, we have constructed an innovative standardized floor plan data structure, which not only achieves consistency and accuracy of data but also greatly improves the efficiency of data processing. The application of this achievement allows designers and researchers to quickly and accurately obtain and analyze key features of interior spaces, laying a solid foundation for subsequent design and optimization work. In addition, the flexibility and scalability of this structure also provide strong support for more complex and changeable interior design needs in the future.

Secondly, the improvement of model interpretability and dynamic adaptability has brought a new perspective to the field of interior design. The selection of the logistic regression algorithm not only enhances the transparency and credibility of the model but also enables designers to better understand and apply the decision logic of the model. More importantly, the dynamic adaptability of the model allows it to self-learn and optimize according to the changes in residents’ needs and the evolution of the environment, achieving continuous evolution of the design. The importance of this adaptability is particularly prominent in the rapidly changing modern living environment. To improve the predictive accuracy and adaptability of the model, this article also carried out innovative feature engineering. By introducing four key features- Space Balance Ratio, Entrance Line Ratio, Kitchen Threshold Distance and Balcony Threshold Distance-and providing detailed calculation formulas, the model can more accurately identify and utilize key features of living spaces. The introduction of these features not only enriches the input dimensions of the model but also provides a deeper spatial perception capability for the model. In terms of model performance, the proposed model in this article has achieved significant results. The TOP5 accuracy rate of the model reached 97.3% and the TOP1 accuracy rate reached 86.4%, indicating that the model has high accuracy and reliability in predicting the layout of living spaces. This achievement not only proves the effectiveness of the model but also provides a new and efficient solution for the field of interior design.

In terms of multi-objective optimization, this article has successfully resolved the conflicts and balance between multiple objectives through simulated annealing algorithms and innovative evaluation functions. This balance is reflected not only in performance indicators but also takes into account many considerations in practical applications, providing the possibility of finding design solutions that perform well at both the theoretical and practical levels. The application of this achievement provides a new way of thinking and method for multi-objective optimization problems in interior design.

In the search for functional areas in geometric space, the ASCRT algorithm strategy proposed in this article demonstrates excellent adaptability and flexibility for irregularly shaped spaces. This algorithm can achieve instant optimization and adjustment of space under multi-dimensional constraints, greatly enhancing the immediacy and flexibility of interior design. The application of this achievement not only solves the limitations of traditional algorithms in the division of irregular spaces but also provides new solutions and ideas for interior design.

In terms of furniture layout strategy analysis, this article deeply analyzes the strategies for furniture layout in the living room, hallway, dining room and corridors, highlighting the core role of furniture placement lines in achieving rational use of space and aesthetic layout. Through careful planning, we have determined suitable placement locations and directions for key furniture such as sofas, TVs, dining tables and shoe cabinets, aiming to optimize space use and enhance the living experience. In strategy formulation, sofas are centered around the focal elements of the living room to enhance spatial cohesion, considering interaction and view. TV placement focuses on viewing comfort, dining tables are conveniently connected to the kitchen and shoe cabinets are easy to use without occupying corridor space. The determination of furniture placement locations and directions integrates spatial layout, functional requirements and aesthetic principles, ensuring spatial fluidity and visual balance.

Meeting personalized needs is another important achievement of this article. By establishing a two-way communication channel between designers and algorithm models, this article not only enhances the level of personalized design but also makes the design process more scientific and efficient. In addition, the various design schemes included in the dataset demonstrate the creativity and ideas of designers, while fully considering the personalized needs of users, providing a more diverse range of solutions for interior design.

This article analyzes the floor plan design of the Binhu City residential complex by Country Garden Group in Hangzhou, China and proposes the application and effect of static panoramic images in interior design display. It intuitively showcases the overall layout of the room, highlighting the functional areas in the living room. The display of these areas not only considers the utilization rate of the space but also takes into account the visual balance, providing a comfortable space for family communication and entertainment.

In summary, this article has achieved significant results in the intelligent division and layout optimization of the living room area. Through standardized data preprocessing, interpretability and dynamic adaptability of models, balance in multi-objective optimization, functional area search in geometric space, furniture layout strategies and satisfaction of personalized needs, this article provides a comprehensive, scientific and practical solution for the field of interior design. These achievements not only promote the technological progress of interior design but also provide strong support for designers and researchers, making a more humanized, intelligent living space possible.

The achievements of this article are not only innovative in theory but also have a wide range of applicability and profound impact in practical applications. Through in-depth exploration and practice in this article, we have gained a deeper understanding of the scientification, intellectualization and personalization of interior design. In the future, we will continue to conduct in-depth research in the following areas:

i. Continue to optimize data standardization processing and feature extraction methods to improve the accuracy and efficiency of data processing.

ii. Explore more advanced machine learning algorithms to enhance the interpretability and dynamic adaptability of models, in order to better adapt to the ever-changing living needs.

iii. Conduct in-depth research and improve multi-objective optimization algorithms to achieve a more comprehensive balance and better design solutions.

iv. Further study geometric space optimization algorithms to improve adaptability and optimization capabilities for complex spatial forms.

v. Expand the dataset by collecting more personalized design schemes to meet the needs of a broader user group.

vi. Engage in deeper collaboration with disciplines such as psychology and sociology to more comprehensively understand and meet the psychological and social needs of residents.

vii. Consider the impact of interior design on environmental sustainability and resource utilization efficiency and promote the concept of green design.

Through these further research and exploration, we believe that the field of interior design will usher in a future that is more intelligent, personalized and scientific. The results of this article will provide a solid theoretical and practical foundation for achieving this goal.           

References

1. Stavre E (2019) The Importance of Architecture and Interior Design in the Albanian Construction Market. Interdiscip J Res Dev 6(1): 58.
2. Birago A (2022) Living Room. KENTE - Cape Coast J Liter Arts 2(2): 1-15.
3. Tadepalli S, Jayasree T, Lakshmi Visakha V, Chelliah S (2021) Influence of ceiling fan induced non-uniform thermal environment on thermal comfort and spatial adaptation in living room seat layout. Building and Environment 205: 108232.
4. Zheng S (2022) An evaluation system for interior design solutions based on artificial intelligence processing technology. Presented at the 7^th International Symposium on Advances in Electrical, Electronics and Computer Engineering (ISAEECE 2022), Xishuangbanna, China.
5. Mitton M, Nystuen C (2007) Residential Interior Design: A Guide to Planning Spaces.
6. Liu CY (2011) The Purpose and Main Basis of House Layout Analysis in Interior Design. Adv Mater Res 250-253: 1113-1117.
7. Saruwono M, Zulkiflin NF, Mohammad NMN (2012) Living in Living Rooms: Furniture Arrangement in Apartment-Type Family Housing. Procedia - Soc Behav Sci 50: 909-919.
8. Qiugang R (2016) Evaluation Research of Interior Design of Living Space Based on Grey System Theory. Int J Hybrid Inform Technol 9(1): 379-386.
9. Liu C, Wu J, Kohli P, Furukawa Y (2017) Raster-to-Vector: Revisiting Floorplan Transformation. Presented at the 2017 IEEE International Conference on Computer Vision (ICCV), Venice.
10. Kán P, Kaufmann H (2017) Automated interior design using a genetic algorithm. Presented at the VRST ’17: 23^rd ACM Symposium on Virtual Reality Software and Technology, Gothenburg Sweden.
11. Uludağ C (2018) Preferences and Placement on Living Room & Dining Room Furniture.
12. Jie Z, Xuejin C (2018) Convertible furniture design. Computers & Graphics 70: 165-175.
13. Yang B, Li L, Song C, Jiang Z, Ling Y (2019) Automatic Furniture Layout Based on Functional Area Division. Presented at the 2019 International Conference on Cyberworlds (CW), Kyoto, Japan.
14. Wang W, Dong S, Zou K, LI W (2020) Room Classification in Floor Plan Recognition. Presented at the ICAIP 2020: 2020 4th International Conference on Advances in Image Processing, Chengdu China.
15. Lv X, Zhao S, Yu X, Zhao B (2021) Residential floor plan recognition and reconstruction. Presented at the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA.
16. Wu W, Feng Y (2022) Interior Space Design and Automatic Layout Method Based on CNN. Math Problem Engineer p. 1-14.
17. Liu Y, He FM (2010) Geomancy in the Living Room and Bedroom Interior Design. Jiangxi University of Architecture and Technology.
18. Jin TL (2014) Comparison of the composition of modern residential communities in South Korea and China (centered around Seoul and Beijing). Tianjin University, Doctoral dissertation.
19. Zhang ZF (2016) A Brief Discussion on the Space Division of Living Room in Interior Design. Popular Literature and Art 9: 2.
20. Zhang Y (2017) Functional Analysis of Interior Design Space Layout. Building Materials and Decoration 47: 1.
21. Wang L (2023) Research on spatial layout and functional optimization in interior design. Footwear Craftsmanship and Design 3(19): 141-143.
22. Yi HS (2023) Research on Space Layout and Functional Zoning Optimization in Architectural Design. Architecture and Construction.
23. Wang YT, Liang C, Huai N, Chen J, Zhang CJ (2023) A Survey of Personalized Interior Design. Computer Graphics Forum. Comput Graph Forum 4(6): e14844.
24. You Y, Liu G, Liu L (2022) Art Design Method of Interior Space Layout Based on CAD Drawing. Computer-Aided Design and Appl 19(S8): 44-54.
25. Yang T, Ibrahim A, Chang-Gyoon L (2018) Analysis of Dimensionality Reduction Methods through Epileptic EEG Feature Selection for Machine Learning in BCI. J Korea Institute Electronic Commun Sci 13(6): 1333-1342.
26. Wei H, Wang X (2024) Financial Risk Management Early-Warning Model for Chinese Enterprises. J Risk Financial Manage 17(7): 255.
27. Kim KC, Hai-tong W (2017) Development of a Face Detection and Recognition System Using a Raspberry Pi. J Korea Institute Electron Commun Sci 12: 859-964.
28. Zheng H, Ren Y (2020) Architectural Layout Design through Simulated Annealing Algorithm. Proceedings of the 25^th Conference on Computer Aided Architectural Design Research in Asia (CAADRIA), Bangkok, Thailand.
29. Adilova A, Shamoi P (2024) Aesthetic Preference Prediction in Interior Design: Fuzzy Approach.
30. Wang YJ (2024) Research and Application of Movie Recommendation Algorithms Based on Knowledge Graph and Graph Attention Network. Fudan University, Master’s thesis, Shanghai, China.