A SYSTEM THAT ENABLES PRODUCTIVE ARTIFICIAL INTELLIGENCE-BASED HAIR AND BEARD DESIGN.

Abstract

This invention relates to a system (1) that enables the production of hair and beard designs by labeling a dataset containing human hair and human beard and training a generative artificial intelligence model with this data.

Description

1 TARIFF PRODUCTIVE AI-BASED HAIR AND BEARD DESIGN A SYSTEM THAT MAKES IT HAPPEN Technical Area 5 This invention involves labeling a dataset containing human hair and human beard, and then analyzing this data. training a generative artificial intelligence model to produce hair and beard designs. It is related to a system that provides. Previous Technique Today, research and development in generative artificial intelligence are ongoing. With the increase in investments, many innovative solutions have been produced. In addition to these solutions Generating different detection and classification models with artificial intelligence models 15 Many studies are ongoing in this area. However, hair and beard design... Studies on solving problems using generative artificial intelligence technologies are widespread. It is not. Therefore, considering the studies and shortcomings in the current technology, 20 when available, the dataset containing human hair and human beard is labeled And with this data, a generative artificial intelligence model can be trained to design hair and beard. It appears that a system is needed to enable its production. Korean patent KR102520839B1, which falls under the prior art, 25 The document discusses how advertising models incorporate artificial intelligence algorithms and virtual hair. A method for placement is described. In the invention in question... in the method, by scanning facial images on a website or application screen The process begins. Then, the information from these facial images is processed into a first-generation generator. By entering the adversarial network (GAN) model, a large number of virtual face images are obtained. 30 The advertising model's face is photographed and numerous virtual faces are created. 2 One of the images is selected. Then the first face image and the selected one are shown. A composite image is obtained by inputting the virtual face image into a second GAN model. The second GAN model is created from face segmentation maps of each. by extracting and determining the hairstyle based on the hair segmentation map It enables the synthesis of hair in the first virtual face image, virtual face 5 It enables the generation of a composite image output by synthesizing it with other materials. The advertising model extracts text from historical data, performs preprocessing, and This involves extracting features related to the advertising model. This method also, including memory and processor, deep GAN models such as the first GAN model and the second GAN model. It provides a GPU device for the use of learning models. Advertisement 10 By determining the model's characteristics such as gender, age, and facial features, a suitable virtual model is selected. It is possible to select facial images. Brief Description of the Invention The aim of this invention is to label a dataset containing human hair and human beard, and Using this data, a generative artificial intelligence model can be trained to design hair and beard. The goal is to implement a system developed to enable its production. Another aim of this invention is to create human skin with different types of hair and beard designs. object tagging format using hair and human beard data, labelimg A labeling process using a library and by manually marking the labels. the process and generative artificial intelligence (GAN, Diffusion) of the labeled data A system developed to provide input to software. to accomplish. 25 Another objective of this invention is to improve the success and stability of the trained artificial intelligence model. face detection and feature extraction to improve and achieve more successful results The goal is to implement a system developed to enable this to happen. 3 Another purpose of this invention is to check whether images contain faces. and obtaining detail of the facial region by performing feature extraction The goal is to implement a system developed to ensure this. Another purpose of this invention is to define the boundaries of the eyes, eyebrows, lips, and cheeks. feature extraction is used to thus identify a human head with hair and beard. By uploading their photos, users can choose the hair and beard style that best suits them. a system developed to enable them to produce the design to accomplish. Another aim of this invention is to provide the user with a visual output of the produced design. provided, and also, on its own image, the final version of this design. developed with the aim of providing a simulation that they can see The goal is to implement the system. Detailed Description of the Invention The "Generative Artificial Intelligence Based" project was carried out to achieve the purpose of this invention. A system that enables hair and beard design is shown in the attached image. and in this form; 20 Figure 1 shows a schematic view of the system that is the subject of the invention. The parts shown in the figure are individually numbered, and these numbers correspond to... The corresponding answers are given below. 25 1. System 2. Electronic Devices 3. Mobile Device 4. Application 30 5. Server 4 6. Processor By labeling the dataset containing human hair and human beard, and with this data... Training a generative artificial intelligence model to generate hair and beard designs. The system that is the subject of the invention was developed to provide (1); 5 - exchanging data using any remote communication protocol to perform and desktop computer and / or portable computer at least one electronic device configured to be such a device (2), - exchanging data using any remote communication protocol to carry out and to be a device in the form of a cell phone 10 at least one configured mobile device (3), - to be run on a mobile device (3), head images including hair and beard to ensure that the images are uploaded and do not contain hair or beard, It is necessary to send an image of a human head with hair and beard. Displaying the warning in this form will show you the most suitable hair and beard design produced. 15 to view, this design's final form is a simulation on the visuals. at least one application configured to display in this way (4), - using any communication protocol with electronic devices (2) to establish a connection, using any communication protocol with a mobile device (3) to connect and through this connection mobile device (3) 20 To communicate with the application (4) that is being run on, application (4) Accessing head images including hair and beard, recorded to hold and transmit, to access the warning mechanism, within the images a human head without hair and beard and a head with hair and beard 25 on the implementation of the warning that the image should be transmitted (4) Presenting the most suitable hair and beard styles for users in the form of notifications. to access the design, to apply this generated design to the user (4) presenting it in visual form and applying this design to the user's own image the simulation application (4) on which the user can see the final version at least one server configured to serve (5), 30 - to be run on electronic device (2) head including hair and beard accessing images, preprocessing images with a generative artificial intelligence model going through the process, including different types of hair and beard designs Object tagging using human hair and human beard data. in format, using the labelimg library and manually marking it 5 This involves performing a labeling process and making the labeled data productive. to be given as input to artificial intelligence (GAN, Diffusion) software to provide, based on GAN-based neural networks and Diffusion models training the artificial intelligence model, developed GAN-based Using Diffusion weighting in the last layer of the neural network, the model's 10 to improve prediction performance, with head images including hair and beard Running a generative artificial intelligence model assesses the success of the trained model and To increase determination and achieve more successful results, facial recognition and Feature extraction involves checking if the image contains a face, hair, and beard. By measuring the quality of the image in this way, the generative artificial intelligence model is 15 It contains at least one processor (6) configured to run. The electronic device (2) in the system (1) which is the subject of the invention, any remote communication to exchange data using the protocol and have at least one The processor (6) is configured to execute. The electronic device (2), desktop 20 It is a device in the form of a computer and / or portable computer. Electronic device (2), any remote communication protocol included in the known state of the art It is configured to connect to the server (5) using. The mobile device (3) included in the system (1) which is the subject of the invention, any remote communication 25 to exchange data using the protocol and have at least one The application (4) is configured to run. Mobile device (3), cell phone It is a device in the form of a mobile device (3), which is included in the known state of the art. connect to the server (5) using any remote communication protocol It is structured accordingly. 30 6 The application (4) in the system (1) which is the subject of the invention, on the mobile device (3) It is structured to be executed. The application (4) has at least one interface. allows users to upload head images including hair and beard. It is structured to provide at least one interface. The application (4) on it users images that do not include hair and beard and hair and 5 the message should include an image of a human head with a beard. It is configured to enable them to view the warning. The application (4), in at least on an interface, users can choose the most suitable hair color produced for them and to allow them to visualize the beard design and to choose the final version of this design themselves. 10 to enable them to view the images in the form of a simulation It is being structured. The server (5) in the system in question (1) is located in the known state of the art. the field connects with electronic devices (2) using any communication protocol to establish and run on electronic device (2) via this established connection 15 The server (5) is configured to communicate with the processor (6). using any communication protocol available in the known state of mobile to connect with device (3) and through this connection mobile device (3) It is configured to communicate with the application (4) being run on it. Server (5) can access head images including hair and beard on application (4), 20 It is configured to keep a record and transmit it to the processor (6). Server (5) accessing the alert mechanism transmitted by the processor (6), hair in the images and a picture of a human head that does not include a beard and has hair and a beard the warning that it should be conveyed in the form of notification upon implementation (4) It is configured to provide. The server (5) provides users 25 via processor (6). To access the most suitable hair and beard design produced for this purpose, and to obtain this produced design... to present the application (4) to the user in a visual form and the user's own (4) It is configured to be presented to the user. 7 The processor (6) located in the system (1) which is the subject of the invention, on the electronic device (2) It is configured to be executed. The processor (6), head containing hair and beard It is configured to access its images. The processor (6) is a generative artificial intelligence. to enable pre-processing of images using the model It is configured. The processor (6) includes 5 different types of hair and beard designs. Using human hair and human beard data in an object tagging format, Labeling can be done using the labelimg library and by manually adding the label. the process is carried out and the labeled data is processed by generative artificial intelligence (GAN, Diffusion) It is configured to provide input to the software. Processor (6), Artificial intelligence based on GAN-based neural networks and Diffusion models 10 It is structured to carry out the training of the model. Processor (6), In the last layer of the developed GAN-based neural networks, Diffusion weighting by using, to improve the model's predictive performance It is configured. The processor (6) produces head images including hair and beard. It is configured to run the artificial intelligence model. The processor (6) is trained on 15 To increase the success and stability of the model, and to achieve more successful results, It is configured to perform face detection and feature extraction. The processor (6), by measuring the quality of the image, such as whether there is a face, hair, and beard in the image. It is configured to run a generative artificial intelligence model. The processor (6), Checking whether the images contain faces and also attribute 20 Extracting the image to obtain detailed information about the face, including the eyes, eyebrows, lips, and cheeks. It is structured to define its boundaries. The processor (6) is shown in the images. If it is determined that the head does not contain hair and beard, then a human head containing hair and beard. to create a warning mechanism that the image needs to be transmitted and It is configured to transmit to the server (5). The processor (6) is configured to transmit the sima-nist service 25 the image included is improved and the improved version is shown to the user. It is designed to function as a sub-service of the Enhance Image service. It is configured. The processor (6) displays images of human heads with hair and beard. by accessing and producing the most suitable hair and beard design for users, these produced to ensure that the design is presented to the user in visual form and the user 30 8 a simulation where he can see the final version of this design on his own image It is structured to create. Industrial application of the invention The invention system includes (1) different types of hair and beard designs for humans. object tagging format using hair and human beard data, labelimg A labeling process using a library and by manually marking the labels. This involves feeding the labeled data into generative artificial intelligence (GAN, Diffusion) software. Thus, the dataset containing human hair and human beard is 10% as input. hair and beard are labeled, and a generative artificial intelligence model is trained using this data. The design and production process is ensured. Around these fundamental concepts, the invention's subject is "Generative AI-Based Hair and..." There are many different 15 related to "A System That Enables Beard Design (1)". It is possible to develop applications, and the invention is illustrated with the examples described here. It cannot be restricted, it is essentially as stated in the claims.

Claims

9 REQUESTS 1. Labeling the dataset containing human hair and human beard, and these data... hair and beard design by training a generative artificial intelligence model. developed to enable its production; 5 - exchanging data using any remote communication protocol to perform and desktop computer and / or portable computer at least one electronic device configured to be such a device (2), - exchanging data using any remote communication protocol to carry out and to be a device in the form of a cell phone 10 at least one configured mobile device (3), - to be run on a mobile device (3), head images including hair and beard to ensure that the images are uploaded and do not contain hair or beard, It is necessary to send an image of a human head with hair and beard. Displaying the warning in this form will show you the most suitable hair and beard design produced. 15 to view, this design's final form is a simulation on the visuals. at least one application configured to display in this way (4), - using any communication protocol with electronic devices (2) to establish a connection, using any communication protocol with a mobile device (3) to connect and through this connection mobile device (3) 20 To communicate with the application (4) that is being run on, application (4) Accessing head images including hair and beard, recorded to hold and transmit, to access the warning mechanism, within the images a human head without hair and beard and a head with hair and beard 25 on the implementation of the warning that the image should be transmitted (4) Presenting the most suitable hair and beard styles for users in the form of notifications. to access the design, to apply this generated design to the user (4) presenting it in visual form and applying this design to the user's own image the simulation application (4) on which the user can see the final version including at least one server (5) configured to serve and; 30 - to be run on electronic device (2) head including hair and beard accessing images, preprocessing images with a generative artificial intelligence model going through the process, including different types of hair and beard designs Object tagging using human hair and human beard data. in format, using the labelimg library and manually marking it 5 This involves performing a labeling process and making the labeled data productive. to be given as input to artificial intelligence (GAN, Diffusion) software to provide, based on GAN-based neural networks and Diffusion models training the artificial intelligence model, developed GAN-based Using Diffusion weighting in the last layer of the neural network, the model's 10 to improve prediction performance, with head images including hair and beard Running a generative artificial intelligence model assesses the success of the trained model and To increase determination and achieve more successful results, facial recognition and Feature extraction involves checking if the image contains a face, hair, and beard. By measuring the quality of the image in this way, the generative artificial intelligence model is 15 characterized by having at least one processor (6) configured to run a system (1).

2. Data exchange using any remote communication protocol. to perform and execute at least one processor (6) on it 20 configured and desktop computer and / or portable computer The claim is characterized by an electronic device (2) which is a device of this type. A system like the one in 1 (1).

3. Connect to the server (5) using any remote communication protocol 25 characterized by an electronic device (2) configured to establish A system like the one in claim 1 or 2 (1).

4. Data exchange using any remote communication protocol. to carry out and implement at least one application (4) on it 30 mobile device (3) which is a device structured and in the form of a cell phone 11 as in any of the above characterized claims system (1).

5. Connect to the server (5) using any remote communication protocol. 5 characterized by mobile devices (3) configured to establish a system like any of the above requests (1).

6. With the application (4) configured to run on the mobile device (3). as in any of the above characterized claims system (1). 10 7. At least one interface should display users' heads including hair and beard. with the application (4) configured to enable them to upload their images as in any of the above characterized claims system (1). 15 8. At least one interface should allow users to see hair and other details within images. a picture of a human head that does not include a beard and includes hair and a beard to enable them to view the warning that it needs to be transmitted 20 of the above requests characterized by the structured application (4). a system like any other (1).

9. At least one interface should be created for users to have the best possible experience. to enable them to view suitable hair and beard designs and this design the final version as a simulation on its own images 25 application configured to provide their viewing (4) as in any of the above characterized claims system (1). 12 10. Using any communication protocol with an electronic device (2) to establish a connection and through this established connection electronic device (2) configured to communicate with the processor (6) running on it any of the above requests characterized by the server (5) such a system (1). 5 11. Connect with mobile device (3) using any communication protocol to establish and through this established connection on the mobile device (3) Server configured to communicate with the running application (4) (5) a 10 as in any of the above claims characterized by system (1).

12. Accessing head images including hair and beard on the application (4), server configured to keep a record and transmit it to the processor (6) (5) like any of the above-mentioned claims characterized by 15 system (1).

13. Accessing the alert mechanism transmitted by the processor (6) within the images a human head without hair and beard and a head with hair and beard 20 on the implementation of the warning that the image should be transmitted (4) characterized by the server (5) configured to present in the form of notification. a system like any of the above-mentioned requests (1).

14. The most suitable hair and beard produced for users via the processor (6). access to the design, this generated design on the application (4) 25 presenting it in visual form and applying this design to the user's own image the simulation application (4) on which the user can see the final version The above is characterized by the server (5) configured to serve a system like any of the requests (1). 13 15. Processor (6) configured to run on electronic device (2) as in any of the above claims characterized by system (1).

16. Processor 5 configured to access head images including hair and beard. (6) like any of the above-mentioned claims characterized by system (1).

17. From the preprocessing step of images with a generative artificial intelligence model 10 characterized by the processor (6) configured to enable the passage. a system like any of the above-mentioned requests (1).

18. Human hair and human beard, featuring different types of hair and beard designs. object labeling format using data, labelimg library a labeling process using and marking by hand 15 the process and generative artificial intelligence (GAN, Diffusion) of the labeled data a processor configured to provide input to its software (6) like any of the above-mentioned claims characterized by system (1).

19. Artificial intelligence based on GAN-based neural networks and diffusion models. with the processor (6) configured to carry out the training of the model as in any of the above characterized claims system (1).

20. In the last layer of the developed GAN-based neural networks, Diffusion weighting by using, to improve the model's predictive performance from the above requests characterized by the configured processor (6) a system like any other (1). 14 21. Generative AI model using head images including hair and beard. characterized by the processor (6) configured to run a system like any of the above requests (1).

22. Increasing the success and commitment of the trained model leads to more successful results. 5 to extract, to perform facial detection and feature extraction from the above requests characterized by the configured processor (6) a system like any other (1).

23. Rate the image quality from 10 to 10 based on whether it contains a face, hair, and beard. configured to run a generative artificial intelligence model by measuring in any of the above requests characterized by processor (6) such a system (1).

24. Checking whether the images contain faces and also, 15 Obtaining detail of the facial area, eyes, eyebrows, by performing feature extraction. with processor (6) configured to define lip and cheek boundaries as in any of the above characterized claims system (1).

25. If it is determined that the images do not contain hair and beard, then hair and beard will be included. a warning that an image of a human head should be transmitted. configured to create the mechanism and transmit it to the server (5) in any of the above requests characterized by processor (6) such a system (1). 25 26. Within the Sima-nist service, the transmitted image is improved and... a sub-service of enhance image that delivers the improved version to the user characterized by the processor (6) configured to serve as a service a system like any of the above requests (1). 30 27. By accessing images of human heads with hair and beards, users can find the best options. to produce a suitable hair and beard design, and to present this design to the user. to provide output in visual form and the user's own image 5. Creating a simulation on which one can see the final version of this design The above is characterized by the processor (6) which is configured to do so. a system like any of the requests (1). 15 25

Images