WO2024220057A1

WO2024220057A1 - Hologram and artificial intelligence supported artificial recognition trainer system

Info

Publication number: WO2024220057A1
Application number: PCT/TR2024/050382
Authority: WO
Inventors: Erdal ÇAYIRCI
Original assignee: Ete Deney Egitim Ve Degerlendirme Teknolojileri AS
Current assignee: Ete Deney Egitim Ve Degerlendirme Teknolojileri AS
Priority date: 2023-04-18
Filing date: 2024-04-07
Publication date: 2024-10-24
Anticipated expiration: 2025-10-18

Abstract

The invention, named as the artificial recognition trainer system, describes a system that will be used in training and exams to recognize subjects and learn their technical features. The invention can also be used for holographic tactical image projection in wargames. Subjects can be any place, terrain, person, event, device, object, vehicle, or any combination of these in any quantity and variety.

Description

HOLOGRAM AND ARTIFICIAL INTELLIGENCE SUPPORTED ARTIFICIAL RECOGNITION TRAINER SYSTEM

Technological Field:

The invention relates to a hologram and artificial intelligence supported training and evaluation system to be used in trainings to recognize the subjects and learn their technical features. The trained subjects can be any place, person, event, device, object or tool. In the following part of the description, the term "artificial recognition trainer" will be used instead of "hologram and artificial intelligence supported artificial recognition trainer".

The State of Art:

Information about the existence of a subject in space arrives in the form of sound or light waves. Holography is a technique that aims to store the information in the waves coming from the subjects in a certain way and to reveal this information again without any loss. The name Holography was given to the technique by Dennis Gabor, who carried out the first studies on this subject. The hologram records the total information of the wave coming from a subject, that is, both amplitude and phase values. When desired, the original wave can be recovered from this recording medium. In the hologram and artificial intelligence supported recognition system, hologram techniques are used to display the image of the subject for which recognition training is given in a three-dimensional space, from the desired angle and including all the details of the subject as possible.

Artificial intelligence is a field of information science that has started to be studied in the second half of the 1950s. In recent years, especially due to the developments in machine learning, new applicable and disruptive artificial intelligence approaches, algorithms and systems in the fields of recognition, understanding and content production are rapidly becoming widespread. In our invention, deep machine learning technologies are used in face recognition, gesture recognition and voice recognition.

The invention, which is subject of the description, describes a system to be used in training and exams to recognize the subjects and learn their technical features. The invention can also be used for holographic tactical picture projection in wargames. Subjects can be any place, terrain, person, event, device, object, vehicle, or any combination of these in any quantity and variety. The system recognizes the faces of people who approach it and are pre-defined in the system, gives them personalized subject recognition training with the voice or gesture-based commands that it receives from these people, and makes an exam when they are ready and evaluates the exam results. During training and exams, trainees and holographic images interact with visual messages given on these images, and artificial voice. The invention, which is subject of the description is about a unique artificial recognition trainer formed by combining advanced holographic imaging and artificial intelligence systems.

In the literature research, it was encountered with the Korean patent numbered KR20210075445. The encountered patent discusses a set of apparatus for editing a hologram video. The hologram and artificial intelligence included in the subject invention and the motivation for educational use are not included in said Korean patent.

Brief Description of the Invention:

The object of the invention is to develop an artificial trainer to be used in subject recognition learning and training. Hologram and artificial intelligence supported artificial recognition trainer can recognize the trainees from their faces, interact with them through voice, gestures, images and messages, and provide autonomously personalized subject recognition training and exams, evaluate the results and rearrange subsequent training sessions autonomously according to the needs of each trainee, based on the performance of the trainees. The artificial recognition trainer can also be used as an interactive holographic sandbox to show the tactical situation in a desired area in wargames.

As the system can work alone, it can also work in centralized modes by connecting to a central computer through computer networks, or in distributed modes by interacting with other artificial recognition trainers through computer networks. Therefore, three working modes are defined:

- Standalone

- Centralized

- Distributed

The artificial recognition trainer is designed to be produced in four types:

- Pyramid projection surface driven by four screens (Figure 2 and Figure 3)

- Pyramid projection surface driven by a screen (Figure 5 and Figure 6)

- Cone projection surface driven by four screens (Figure 2 and Figure 3)

- Cone projection surface driven by a screen (Figure 5 and Figure 6)

The technical features of the artificial recognition trainer are described by using the drawings defined below.

Description of the Drawings:

The invention will be described with reference to the attached drawings, so that the characteristics of the invention will be understood more clearly. However, this is not intended to limit the invention to these particular embodiments. On the contrary, it is intended to cover all alternatives, modifications and equivalences that can be included within the scope of the invention as defined by the attached claims. It should be understood that the details shown are only for illustrating preferred embodiments of the present invention and are intended to provide the most useful and easily understandable description of both the embodiment of the methods and the rules and conceptual features of the invention. In these drawings;

Figure - 1 A schematic view showing the general system architecture.

Figure - 2 A schematic view showing the pyramid projection surface design driven by four screens from the front.

Figure - 3 A schematic view showing the pyramid projection surface design driven by four screens from the top.

Figure - 4 A schematic view showing the pyramid from the top in the pyramid projection surface design driven by four screens.

Figure - 5 A schematic view showing the pyramid projection surface design driven by a screen from the front.

Figure - 6 A schematic view showing the pyramid projection surface design driven by a screen from the top.

Figure - 7 A schematic view showing the pyramid from the top in a pyramid projection surface design driven by a screen.

The side and back views of the designs are the same as the front views. However, there are cameras, speakers and microphones on the sides and back if desired. In the cone design, the only difference is that the pyramid is replaced by a cone. The drawings that will help understand this invention are numbered as indicated in the attached drawings and are given below with their names.

Description of References:

1. Main unit

2. Screen

3. Hologram projection surface

4. Control unit

5. Control computer or processor unit 6. Computer network

7. Camera

8. Microphone

9. Speaker

Description of the Invention:

In this detailed description, the artificial recognition trainer system, which is the subject of the invention, is described only for a better understanding of the subject, with examples that will not form any limiting effect. In the description, it is described a hologram and artificial intelligence supported artificial recognition trainer that will be used in training fast and accurate identification of subjects. The subjects that the system will train can be any place, terrain, person, event, device, object, vehicle or a mixture of all these. The artificial recognition trainer system can also be used as an interactive holographic sandbox to show the tactical situation in a desired area in wargames.

As seen in Figure 1 , the artificial recognition trainer system consists of five main components placed in or around a main unit (1 ). The first component, the screen (2), can be a standard television, monitor, smartphone or tablet, or it can be a screen (2) specially developed for the artificial recognition trainer. The screen (2) reflects images that may contain various subjects onto the hologram projection surface (3). The hologram projection surface (3) can be in pyramid form or cone form. When it is in pyramid form, the hologram images projected on it are reflected to the audience from four different angles. On the hologram projection surface (3), which is in conical form, the subject is shown from the same angle, no matter which angle the viewer watches from. In the artificial recognition trainer system, which is subject of the invention, there is a control unit (4) used to perform the operations of zooming out, zooming in and rotating the images in all directions on the hologram projection surface (3). The images transferred to the hologram projection surface (3) by the screen (2) or screens (2) are transmitted to the screens (2) by a control computer or processor unit (5) and changed depending on user demand. The control computer or processor unit (5) may be a personal computer that can be connected to the computer network (6) and runs on Linux or Windows operating systems, or it may be a processor unit specially developed for the artificial recognition trainer. The computer network (6) is a local or wide area network to provide data and command transfer between the displays (2), the control unit (4), the control computer or the processor unit (5).

There are various alternative models in the artificial intelligence trainer system, which is subject of the invention. For example, the system may comprise one screen (2) or four screens (2). For this reason, it is defined in the description as comprising at least one screen (2). A hologram projection surface (3) in conical form or pyramid form can be obtained with one screen (2), or a hologram projection surface (3) in conical form or pyramid form can be obtained with four screens (2). In Figure 2 and Figure 3, the view of creating a pyramid-shaped hologram projection surface (3) with four screens (2) is given. In Figure 4 and Figure 7, the hologram projection surface (3) views created in pyramid form are given.

In the invention, there is at least one microphone (8) on the screen (2) or on the main unit (1 ) to detect the user's voice command or environment sounds, at least one camera (7) that can perform the user's facial recognition and take images from the environment, and at least one speaker (9) that can perform sound outputs.

There are two modes of the system, which is subject of the invention. These are training mode and evaluation mode. The system operates in these modes as follows:

Initialization and authentication: When a person who has already been introduced to the system sits in front of the system and waits for S seconds, the system recognizes the person by means of the artificial intelligence- based facial recognition and asks the person what he/she wants to do with a message displayed on the holographic projection surface (3) or an artificial voice output from the speaker (9). The S seconds to wait in front of the system is a system setting parameter and can be changed by the system administrator. The user is given at least three options at this stage: “Training”, “Evaluation”, “Exit”. The identified person selects one of these options by voice command or keyboard command via the control unit (4). If one of the “Training” or “Evaluation” options is selected, the system starts the training or evaluation process. If the “Exit” option is selected, the system stops. The person can give the "Exit" command at any time after starting the training or evaluation process. In the voice commands given, the system compares the person's voice with the voice defined in the system. If there is no match, it does not accept the commands and answers given.

Training or evaluation: The system works almost the same way in training and evaluation modes. The main differences between the two modes are as follows:

- In the training mode, the system continues training until the person in training gives the "Exit" command or until all subjects in the system's database that belongs to the selected category are completed. In the evaluation mode, the system runs and stops for only D subjects. The D number is a system setting parameter and can be changed by the system administrator.

- In the training mode, the person being trained can choose the category in which he/she wants to receive training by giving commands such as "Marine Vehicles", "Land Vehicles". In the evaluation mode, subjects are randomly selected by the artificial recognition trainer.

- In the evaluation mode, the system stores data about the answers given by the trainee in a database, such as the time it takes to give the answer and the accuracy of the answer, according to this data, the evaluation results can be shared with the person being evaluated and again stored in the database. During training and evaluation, the system randomly selects a subject, and displays the holographic image of the subject again from a randomly selected angle as the furthest, in other words smallest, on the hologram projection surface (3) and begins to zoom the image at a certain speed. As soon as the user recognizes the object, he/she gives the system a voice or gesture command "Stop" or says the name of the subject. The system stores data such as the time it takes to identify the subject and the size of the subject at the time it is identified and adds it to the evaluation results. If the identification is performed correctly, the system asks the trained/evaluated person randomly selected A/ questions about the subject's technical characteristics and receives the answers via voice or gesture commands. The number A/ is a system setting parameter and can be changed by the system administrator. While in the training mode, the trainee can rotate, minimize or enlarge the subject with voice or gesture commands. If desired, images and features of more than one system can be compared.

Claims

1. The invention is a hologram and artificial intelligence supported training and evaluation system, characterized in that, it comprises the following:

• at least one screen (2) capable of projecting images that may contain various subjects as holograms,

• at least one hologram projection surface (3) that enables the images containing subjects that belong to the trainings projected from said screen (2) to appear as a hologram or three-dimensional,

• a control unit (4) used to perform the operations of zooming out, zooming in and rotating the images in all directions on the hologram projection surface (3),

• at least one control computer or processor unit (5) that transmits the images transferred to the hologram projection surface (3) by the screen (2) or screens (2) to the screens (2) according to the commands received from the user,

• at least one computer network (6) that may be a local or wide area network to provide data and command transfer between displays (2), control unit (4), control computer or processor unit (5).

2. An artificial recognition trainer system according to claim 1 , wherein; it comprises at least one camera (7) that allows the user to be recognized by the system by scanning his/her face and can take images from the environment.

3. An artificial recognition trainer system according to claim 1 , wherein; it comprises at least one microphone (8) that allows the user's voice commands to be received and processed by the system as voice commands, and receives the environment sounds.

4. An artificial recognition trainer system according to claim 1 , wherein; it comprises at least one speaker (9) that provides sound output from the system to the environment and gives voice notifications to the user.

5. An artificial recognition trainer system according to any of the claims above, wherein; it comprises at least one control computer or processor unit (5) that understands the commands such as "Next", "Previous", "Pass", "Zoom In", "Zoom Out", "Rotate" given by the users with their signs and/or gestures in the images taken through the system's camera (6) by means of the artificial intelligence, and capable of running in accordance with these commands.

6. An artificial recognition trainer system according to any of the claims above, wherein; it comprises at least one control computer or processor unit (5) that can recognize the voice in the commands and answers it receives through the system's microphone (7) by means of the artificial intelligence and compare it with the identity of the person it recognizes through the image and if theincompatibility do not match, it may not respond to the commands it receives.

7. An artificial recognition trainer system according to any of the claims above, wherein; it comprises at least one system database containing geographical data of the subjects, three-dimensional and animated subject models of the images to be transferred onto the hologram projection surface (3), and where the system outputs, user data, commands and results are stored.

8. An artificial recognition trainer system according to any of the claims above, wherein; it comprises at least one control computer or processor unit (5) that demonstrates the user identity by comparing the system users' data preloaded into the system database with the data received from the microphone (8) and camera (7) hardware.

9. An artificial recognition trainer system according to any of the claims above, wherein; it comprises at least one control computer or processor unit (5) that demonstrates the user identity by comparing the system users' data preloaded into the system database with the data received from the microphone (8) and camera (7) hardware.

10. An artificial recognition trainer system according to any of the claims above, wherein; it comprises at least one system database where the previous training and evaluation data of the trained people are stored in order to provide personalized training to its users by analyzing these data.

11. An artificial recognition trainer system according to any of the claims above, wherein; it comprises at least one control computer or processor unit (5) that analyzes the data in the database where the system evaluation results are stored, decides whether the trainees have reached the desired level, and ranks them according to these results.

12. An artificial recognition trainer system according to claim 1 , wherein; it comprises at least one control computer or processor unit (5),

• central by connecting via the computer network (6) and interacting with a central system,

• distributed by interacting with other artificial recognition trainers through the computer network (6),

• or that can work alone using the existing database.

13. An artificial recognition trainer system according to claim 1 , wherein; it comprises at least one main unit (1 ) to maintain said screen (2), control unit (4), control computer or processor (5), camera (7), microphone (8) and speaker (9) on and around it.

14. An artificial recognition trainer system according to claim 1 , wherein; the hologram projection surface (3) has a conical shape to ensure that it is shown from the same angle no matter which angle the viewer watches.

15. An artificial recognition trainer system according to claim 1 , wherein; the hologram projection surface (3) has a pyramid form to enable it to be displayed from four different angles on its four surfaces.