KR102533499B1 - User-customized interior design recommedating platform based on artificial intelligence - Google Patents

Abstract

The present invention relates to an artificial intelligence-based customized interior design recommendation platform. The interior design recommendation platform comprises: a management server that provides a user-customized interior design recommendation service; and a user terminal connected to the management server through a network. Therefore, the interior design recommendation platform can increase convenience, satisfaction and the like with interior design.

Description

AI-based customized interior design recommendation platform {USER-CUSTOMIZED INTERIOR DESIGN RECOMMEDATING PLATFORM BASED ON ARTIFICIAL INTELLIGENCE}

The present invention relates to an AI-based customized interior design recommendation platform.

With the development of new technology, the increase in single-person households, and the prolonged COVID-19, home activities are increasing, and demand for small-scale remodeling and self-interiors is increasing. In addition, the number of uses of large furniture specialty stores, DIY specialty stores, and online interior platforms is increasing. In particular, the online interior platform, which is showing a remarkable rise, is expanding its service by increasing the number of users, led by various start-ups.

The space of a house is not just a space to rest and handle private matters, but a specific space in the house is changing into a space where you can do home work, and as you naturally spend more time at home, the existing concept of the structure and interior of the house has changed. is being corrected

However, the online interior platforms currently in service only provide limited services such as construction cases, making it difficult to apply designs suitable for increasingly diverse people's preferences and target spaces. In particular, for interior changes, users (customers) have no choice but to receive limited services from nearby interior companies. There are unreasonable factors such as the need to directly measure the actual dimensions of the space, the first consultation being charged, and the need to book a separate visit to the house to receive consultation service, which takes a considerable amount of time.

In addition, in the case of individuals who construct self-interiors, there is a disadvantage that a lot of time and goods are wasted due to trial and error, and in the end, various technologies, auxiliary materials, tools, etc. must be prepared, and there is a possibility that the construction will fail.

Therefore, a personalized interior coordination function is required for each user, and a demand for a technology capable of quickly changing the concept of space and freely applying recommended interior elements is increasing.

One aspect of the present invention discloses a platform that receives interior data from a user terminal and provides a design image matching the interior data as a recommendation list.

Another aspect of the present invention discloses a platform that recommends a customer-customized interior design based on artificial intelligence learned by converting images that consumers click online into big data.

The technical problem of the present invention is not limited to the technical problem mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art from the following description.

The platform of the present invention includes a management server providing a user-customized interior design recommendation service; and a user terminal connected to the management server through a network.

On the other hand, the management server may include: a member information collection unit for collecting member information including interior design keywords and image click information through the user terminal through the user terminal; an interior design storage unit that acquires an interior design image from an external server, matches keywords to the acquired interior design image, and generates a stored database; a member information analyzer configured to analyze the member information and extract a member-customized interior design image from the database; and an interior design recommendation unit that maps the interior design image extracted by the member information analysis unit to a virtual house space and outputs the image through the user terminal.

Meanwhile, the interior design storage unit receives interior design images from an external server in a first cycle and stores the interior design images in the database, extracts previously stored interior-related texts by crawling SNS websites, and determines that the number of interior-related texts exceeds a threshold value. When extracted, updating the database by receiving an interior design image from an external server at a second period shorter than the first period and storing it in the database;

The member information analysis unit acquires a plurality of interior design images as training data, obtains keywords as labels for each interior design, generates an artificial neural network that has learned the training data and the labels, and sends the artificial neural network to the artificial neural network. You can extract member-customized interior design keywords by entering image click information.

According to one aspect of the present invention described above, a user-customized interior design image is recommended, and a user's activity is analyzed through an artificial intelligence neural network to predict an interior design concept preferred by the user and a user-customized interior design image is recommended. Convenience and satisfaction can be improved in interior design.

1 is a conceptual diagram of an artificial intelligence-based customized interior design platform according to an embodiment of the present invention.
FIG. 2 is a control block diagram of the management server shown in FIG. 1 .
3 and 4 are diagrams for explaining another embodiment of a deep learning operation method using an artificial neural network.

Advantages and features of the present invention, and methods of achieving them, will become clear with reference to the detailed description of the following embodiments taken in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but will be implemented in various different forms, only these embodiments make the disclosure of the present invention complete, and common knowledge in the art to which the present invention belongs. It is provided to fully inform the holder of the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numbers designate like elements throughout the specification.

Terminology used herein is for describing the embodiments and is not intended to limit the present invention. In this specification, singular forms also include plural forms unless specifically stated otherwise in a phrase. As used herein, “comprises” and/or “comprising” do not preclude the presence or addition of one or more other components, steps, and operations to the stated components, steps, and operations.

1 is a conceptual diagram of an artificial intelligence-based customized interior design platform according to an embodiment of the present invention.

Referring to FIG. 1 , a platform 1 according to an embodiment of the present invention may include a management server 10 and a user terminal 20 .

The management server 10 has the same configuration as a conventional web server or web application server in terms of hardware. However, in terms of software, it includes program modules implemented in various languages such as C, C++, Java, PHP, .Net, Python, and Ruby to perform various functions.

The management server 10 may be an operation server of a platform for providing an interior design recommendation service. The management server 10 may manage member information subscribed to the platform and provide a user-customized interior design recommendation service.

The user terminal 20 may include a general PC such as a general desktop or laptop computer, as well as mobile terminals such as smart phones, tablet PCs, PDAs (Personal Digital Assistants), and mobile communication terminals. It can be interpreted as any device capable of web browsing function.

The user terminal 20 may have a browser installed that works with an operating system. This browser may be a basic browser basically installed together with the operating system or a browser additionally installed by a user separately from the basic browser.

The operating system installed on the user terminal 20 is an operating system such as Windows or Macintosh installed on a general PC such as a desktop or Android or iOS installed on a mobile terminal such as a smartphone or tablet PC. It may also be a mobile operating system such as

The management server 10 may be generally connected to the user terminal 20 through a network such as an open computer network such as the Internet. The network is a network that is a transmission path of the web page, and may be a closed network such as a LAN (Local Area Network) or a WAN (Wide Area Network), but is preferably an open type network such as the Internet. The Internet is based on the TCP/IP protocol and several services that exist on its upper layer, such as HTTP (HyperText Transfer Protocol), Telnet, FTP (File Transfer Protocol), DNS (Domain Name System), SMTP (Simple Mail Transfer Protocol), SNMP ( Simple Network Management Protocol), Network File Service (NFS), and Network Information Service (NIS). In addition, when the user terminal 20 is a mobile communication terminal or a smart phone, the network may further include a mobile communication network or a Wi-Fi network.

FIG. 2 is a control block diagram of the management server shown in FIG. 1 .

Referring to FIG. 2 , the management server 10 may include a member information collection unit 11 , an interior design storage unit 12 , a member information analysis unit 13 and an interior design recommendation unit 14 .

The member information collection unit 11 may receive member information through the user terminal 20 .

Member information may include interior design keywords, image click information through the user terminal 20 , and the like.

The interior design storage unit 12 may acquire and store an interior design image from an external server.

The interior design storage unit 12 may create an interior design image database in which keywords are matched and stored with interior design images.

The interior design storage unit 12 may update the interior design image database at predetermined intervals. For example, the interior design storage unit 12 may receive interior design images from an external server in a first cycle and store them in an interior design image database. The interior design storage unit 12 crawls websites such as SNS to extract interior-related text stored in advance, and when the number of interior-related texts is extracted beyond a threshold value, the interior design image database with a second period shorter than the first period. can be updated. That is, the interior design storage unit 12 secures various and up-to-date interior design images by adjusting the update cycle of the interior design image database according to the latest interior issues.

The member information analysis unit 13 may analyze the member information and extract a member-customized interior design image from the database.

For example, the member information analyzer 13 may extract an interior design image from a database as an interior design image customized for a member by using an interior design keyword included in member information.

The member information analysis unit 13 analyzes image click information included in member information through an artificial intelligence neural network to extract interior design keywords, and extracts interior design images from the database as member-customized interior design images using the interior design keywords. can do. The image click information may be an image clicked on a website such as SNS through the user terminal 20 .

For example, the member information analysis unit 13 may extract interior design keywords from image click information using a deep learning-based artificial intelligence learning algorithm.

The member information analysis unit 13 may obtain training data and labels. The member information analyzer 13 may acquire a plurality of interior design images as training data and obtain keywords as labels for each interior design image.

The member information analysis unit 13 may generate an input of an artificial neural network from training data. The member information analysis unit 13 may use the training data as input to the artificial neural network as it is or may generate an input of the artificial neural network after a known process such as removing unnecessary information from each training data.

The member information analysis unit 13 may apply the input to the artificial neural network. Here, the artificial neural network may be an artificial neural network that is learned according to supervised learning. The artificial neural network may be a convolutional neural network (CNN) or a recurrent neural network (RNN) structure suitable for learning through supervised learning.

The member information analysis unit 13 may obtain an output from an artificial neural network. An output of the artificial neural network may be a keyword according to image click information.

The member information analyzer 13 may compare the output of the artificial neural network and the label. The process of comparing the output of the artificial neural network corresponding to inference and the label corresponding to the correct answer may be performed by calculating a loss function. As the loss function, a known mean squared error (MSE), cross entropy error (CEE), or the like may be used. However, the present invention is not limited thereto, and loss functions used in various artificial neural network models may be used if the deviation, error, or difference between the output of the artificial neural network and the label can be measured.

The member information analysis unit 13 may optimize the artificial neural network based on the comparison value. By updating the weights of the nodes of the artificial neural network so that the learning device comparison value gradually decreases, the output of the artificial neural network corresponding to inference and the label corresponding to the correct answer can be gradually matched, and through this, the artificial neural network can It can be optimized to output inferences that are close to the correct answer. For example, the member information analyzer 13 may optimize the artificial neural network by repeating a process of resetting the weight of the artificial neural network so that the loss function corresponding to the comparison value approaches the estimated minimum value. A well-known backpropagation algorithm, stochastic gradient descent, or the like may be used to optimize the artificial neural network. However, it is not limited thereto, and weight optimization algorithms used in various neural network models may be used. The member information analysis unit 13 may train the artificial neural network by repeating the above process.

The member information analyzer 13 may output keywords by inputting image click information received from the user terminal 20 to the artificial neural network.

The member information analysis unit 13 may output an interior design image from the database with the output keyword.

3 and 4 are diagrams for explaining another embodiment of a deep learning operation method using an artificial neural network.

Artificial intelligence (AI) algorithms including deep learning input input data into an artificial neural network (ANN), learn output data 30 through operations such as convolution, and learn Features can be extracted using an artificial neural network. An artificial neural network may refer to a computer scientific architecture modeling a biological brain. In an artificial neural network, nodes corresponding to brain neurons are connected to each other and collectively operate to process input data. Examples of various types of neural networks include Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), Deep Belief Networks (DBNs), and Restricted Boltzman Machines (Restricted Boltzman Machines). Machine, RBM) method, etc., but is not limited thereto. In a feed-forward neural network, neurons in the neural network have links to other neurons. Such connections may extend through the neural network in one direction, for example, in a forward direction.

It shows the structure of an artificial neural network (for example, a Convolution Neural Network (CNN) 20) that receives input data and outputs output data. The artificial neural network may be a deep neural network having two or more layers.

A convolutional neural network can be used to extract “features” such as borders, line colors, etc. from input data. A convolutional neural network may include a plurality of layers. Each layer may receive data, and may process data input to the corresponding layer to generate data output from the corresponding layer. Data output from the layer may be a feature map generated by performing a convolution operation on an image or a feature map input to the convolutional neural network and a weight value of a filter. Initial layers of a convolutional neural network can be operated to extract low-level features such as edges or gradients from the input. Subsequent layers of the convolutional neural network can extract progressively more complex features within the image.

A convolutional neural network may generate a trained neural network by repeatedly training (learning) a given initial neural network. Creating a trained neural network can mean determining neural network parameters. Here, the parameters may include, for example, various types of data input/output to the neural network, such as input/output activations, weights, and biases of the neural network. As the iterative training of the neural network progresses, the parameters of the neural network can be tuned to compute a more accurate output for a given input.

A trained neural network according to an embodiment may include a plurality of neural networks. For example, the trained neural network may include a first neural network and a second neural network. The first neural network and the second neural network may be independently learned or may be learned in association with each other.

The interior design recommendation unit 14 may extract a plurality of interior design images through the member information analysis unit 13 , list the extracted interior design images, and transmit them to the user terminal 20 .

The interior design recommendation unit 14 may provide a service of 3D mapping the extracted interior design image to the virtual house space and outputting the result.

The platform 1 according to an embodiment of the present invention like this recommends a user-customized interior design image, further analyzes the user's activity through an artificial intelligence neural network, predicts the user's preferred interior design concept, and user-customized interior design. By recommending design images, convenience and satisfaction in interior design can be increased.

5 is a diagram showing an output device according to an embodiment of the present invention.

Referring to FIG. 5 , the output device 30 may include an output plate 31 , a connecting member 32 and a lighting member 33 .

The output plate 31 may be installed on a wall. The output plate 31 may be applied as a hologram film for displaying a three-dimensional stereoscopic image.

The connection member 32 may be installed on a wall surface of the output plate 31 . For example, the connection member 32 may be provided in the form of a tube in which wires can be embedded, and a combination of one or more straight or curved fittings to maintain a certain distance from the output plate 31. can be provided with

The lighting member 33 may be installed on the connecting member 32 to emit light to the output plate 31 . For example, the lighting member 33 may be provided with a halogen lamp, a natural light fluorescent lamp, or the like.

Such an output device 30 may be connected to the management server 10 through a network, receive hologram information of an interior design image, and output the hologram information to the output plate 31 .

Meanwhile, the lighting member 33 may include a side bar 35, a connection frame 36, and a suction pipe 37.

The side bar 35 may be installed along the outer circumferential surface of the lighting member 33 . At this time, the side bar 35 may be installed to be rotatable with respect to the outer circumferential surface of the lighting member 33 .

The connection frame 36 may be formed in a tubular shape and may be formed through the side bar 35 .

The suction pipe 37 may be formed in a pipe shape, and one end may be installed in the connection frame 36 and the other end may be installed in the suction module.

Such a lighting member 33 can increase visibility by absorbing foreign substances in a light emitting direction of the lighting member 33 through the suction pipe 37 at a predetermined period. In addition, the angle of the side bar 35 can be adjusted by a user's manipulation.

Although the embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art can realize that the present invention can be implemented in other specific forms without changing the technical spirit or essential features. you will be able to understand Therefore, the embodiments described above should be understood as illustrative in all respects and not limiting.

1: Interior design recommendation platform
10: management server
20: user terminal

Claims (3)

A management server that provides a user-customized interior design recommendation service; and
Including; a user terminal connected to the management server through a network;
The management server,
a member information collection unit that collects member information including interior design keywords and image click information through the user terminal through the user terminal;
an interior design storage unit that acquires an interior design image from an external server, matches keywords to the acquired interior design image, and generates a stored database;
a member information analyzer configured to analyze the member information and extract a member-customized interior design image from the database; and
An interior design recommendation unit for mapping the interior design image extracted by the member information analysis unit to a virtual house space and outputting the image through the user terminal;
The interior design storage unit,
In the first cycle, interior design images are received from an external server and stored in the database, and SNS websites are crawled to extract interior-related text stored in advance. Updating the database by receiving an interior design image from an external server at a short second cycle and storing it in the database;
The member information analysis unit,
Acquiring a plurality of interior design images as training data, acquiring keywords as labels for each interior design, generating an artificial neural network that has learned the training data and the labels, and inputting the image click information into the artificial neural network Extract member-customized interior design keywords,
An output device connected to the management server through a network to receive and output hologram information of an interior design image;
The output device is
an output plate installed on a wall, applied with a hologram film for displaying a three-dimensional stereoscopic image, and outputting hologram information received from the management server;
a connecting member formed in a tubular shape into which wires can be embedded and installed on a wall surface on which the output plate is installed; and
Including; a lighting member installed on the connecting member to emit light to the output plate;
The lighting member,
Side bars installed along the outer circumferential surface of the lighting member and rotatably installed with respect to the outer circumferential surface of the lighting member;
A connection frame formed in a tubular shape and penetrating the side bar; and
A suction tube formed in a tube shape, one end installed in the connection frame and the other end installed in the suction module to absorb foreign substances in the light emitting direction of the lighting member at a predetermined cycle; interior design recommendation platform.
delete delete

Images