KR20240030033A - System and method for actor casting based on artificial intelligence technology - Google Patents

Abstract

The present invention relates to actor recommendation technology, and more specifically, to an artificial intelligence-based actor casting system that analyzes users' actor evaluation information to generate actor image keywords and recommends audition/casting officials the most suitable actor for the role. and methods. To this end, the artificial intelligence-based actor casting platform according to the present invention receives gender, age, physical information, character/image information, and role information as setting information for actor casting from a casting person's terminal, extracts feature vectors from each information, and extracts feature vectors from each information. It includes an actor recommendation unit that analyzes feature vectors through a neural network and outputs casting information, and an actor evaluation unit that analyzes actor evaluation information about already registered actors collected from the user terminal and outputs image keywords.

Description

{System and method for actor casting based on artificial intelligence technology}

The present invention relates to actor recommendation technology, and more specifically, to an artificial intelligence-based actor casting system that analyzes users' actor evaluation information to generate actor image keywords and recommends audition/casting officials the most suitable actor for the role. and methods.

While there are more than 10,000 auditions for roles per year, there are more than 15 million audition applications for actors per year, resulting in a very severe information asymmetry in the audition/casting market.

Except for famous actors, it is difficult for most ordinary actors to know audition information and audition application results, and production companies are having a lot of trouble finding suitable actors among tens of thousands of actors.

As such, both actors and production companies are currently wasting a lot of time and money on auditions/casting.

Meanwhile, recently, the technology called machine learning or machine learning has been applied to various fields ranging from software technology to finance and economics, and has become a core technology leading the rapid development in the fields of computer vision and image processing in particular.

This machine learning technology is a field of artificial intelligence (AI) and refers to algorithms and related fields that enable analysis of new data by learning patterns or characteristics from given data. In particular, as machine learning techniques called deep learning emerge as core technologies, interest in related technologies and application fields is gradually increasing.

The deep learning technique is a model of an artificial neural network that mimics the nervous system of a living organism. While the existing artificial neural network model consists of connections between thin layers of neuron models, the deep learning technique builds up layers of neuron models deeply. This is a technology that applies a model that increases the learning ability of a neural network by increasing it.

The concept of deep learning as an artificial neural network made up of multiple layers was proposed in the 1970s, but stagnated due to the complexity of learning calculations. However, its performance has been improved through several recent studies, and related studies have been conducted on areas such as voice recognition and image recognition. The demand is growing rapidly as the field shows outstanding results.

Conventionally, an actor matching service is provided on an online platform by storing the actor's profile information in a database, but this is only simple matching and does not sufficiently satisfy the needs of audition/casting officials.

Accordingly, there is a need for a new service that can quickly and accurately match audition/casting officials with the optimal actor for the role by applying artificial neural networks, which are applied in various fields, to the conventional audition/casting platform.

Korean Patent Publication No. 2021-0135943

The present invention was created to solve the above problems, and the purpose of the present invention is to quickly and accurately recommend the optimal actor for the role to audition/casting officials.

Another purpose of the present invention is to enable actors to have their images objectively evaluated by the public.

To this end, the artificial intelligence-based actor casting platform according to the present invention receives gender, age, physical information, character/image information, and role information as setting information for actor casting from a casting person's terminal, extracts feature vectors from each information, and extracts feature vectors from each information. It includes an actor recommendation unit that analyzes feature vectors through a neural network and outputs casting information, and an actor evaluation unit that analyzes actor evaluation information about already registered actors collected from the user terminal and outputs image keywords.

Here, the neural network of the actor recommendation unit is characterized by machine learning using learning data that uses gender, age, physical information, character/image information, and role information as input data and casting information consisting of top-ranked actors as output data. do.

Additionally, the actor evaluation unit is characterized by analyzing actor evaluation information using an unsupervised learning model.

The artificial intelligence-based actor casting method according to the present invention is an actor casting method performed on an artificial intelligence-based actor casting platform. Actor evaluation information for already registered actors is collected from the user terminal and analyzed using an unsupervised learning model for each actor. A step of outputting an image keyword for an actor, using gender, age, physical information, character/image information based on the image keyword, and role information as input data, and casting information consisting of at least one actor among the already registered actors. A step of learning an actor recommendation model using output data, and receiving gender, age, physical information, character/image information, and role information as setting information for actor casting from the casting personnel terminal, the learned actor recommendation model is generated. It includes a step of analyzing and outputting casting information.

As described above, according to the present invention, images of actors evaluated by general users are collected and image keywords that the actor is projected to people are generated and displayed, so actors can confirm a more objective image that the public has of themselves. It works.

Additionally, casting officials can find the optimal actor for the role simply by entering the desired actor's physical information, character/image information, and role information.

1 is a schematic configuration diagram of an artificial intelligence-based actor casting system according to the present invention.
Figure 2 is a signal flow diagram between each subject in the artificial intelligence-based actor casting system according to the present invention.
Figure 3 is an internal configuration diagram of an artificial intelligence-based actor casting platform according to the present invention.
Figure 4 is an internal configuration diagram of an actor evaluation unit according to the present invention.
Figure 5 is a flowchart of the artificial intelligence-based actor casting method according to the present invention.

Below, with reference to the attached drawings, embodiments of the present invention will be described in detail so that those skilled in the art can easily implement the present invention. However, the present invention may be implemented in various different forms and is not limited to the embodiments described herein. In order to clearly explain the present invention in the drawings, parts that are not related to the description are omitted, and similar parts are given similar reference numerals throughout the specification.

Throughout the specification, when a part is said to “include” a certain component, this means that it may further include other components rather than excluding other components, unless specifically stated to the contrary.

In addition, the terms “…unit” and “…module” used in the specification refer to a unit that processes at least one function or operation, which may be implemented as hardware, software, or a combination of hardware and software.

Hereinafter, an artificial intelligence-based actor casting system and method according to an embodiment of the present invention will be described in detail with reference to the drawings.

Figure 1 shows a schematic configuration of an artificial intelligence-based actor casting system according to the present invention.

Referring to FIG. 1, the artificial intelligence-based actor casting system according to the present invention includes an actor casting platform 100 and a plurality of computing devices 200.

The actor casting platform 100 is a web-based or app-based platform that can be accessed through a network by a plurality of computing devices 200. The actor casting platform 100 is accessed by a web browser or client application installed on the computing devices 200. The actor casting platform 100 may be implemented on a server built by an actor casting service company.

The actor casting platform 100 receives setting information for actor casting from audition/casting officials and performs an actor recommendation process. Additionally, the actor casting platform 100 collects actor-related evaluation information from general users and performs an actor image evaluation process.

The plurality of computing devices 200 connected to the actor casting platform 100 are computing devices owned by actor casting officials, general users, actors, etc. The computing device may be a smartphone, tablet PC, laptop computer, personal computer, etc., and any type of device may be used as long as it can access the actor casting platform 100 through an app or web.

For convenience of explanation, the plurality of computing devices 200 include an audition/casting official terminal 200-1 and a general user terminal 200-2, respectively, a casting official terminal 200-1 and a user terminal ( It is abbreviated as 200-2).

The casting person terminal 200-1 is a computing device owned by a company or person interested in casting an actor, and the user terminal 200-2 is a computing device owned by a general user or actor interested in an actor.

In this way, when the casting official terminal 200-1 accesses the actor casting platform 100 and inputs the setting information required for actor casting, the actor casting platform 100 analyzes the entered setting information using an artificial intelligence-based model. to output casting information.

Additionally, when the user terminal 200-2 accesses the actor casting platform 100 and inputs image evaluation information for a registered actor, the actor casting platform 100 collects the image evaluation information and uses an artificial intelligence-based model. Analyze and output image keywords.

Figure 2 shows the signal processing flow between each subject in the artificial intelligence-based actor casting system according to an embodiment of the present invention.

In order to provide an artificial intelligence-based actor casting service according to the present invention, an actor pool and an artificial intelligence-based analysis model that has been trained using the actor casting platform 100 must be established.

Referring to FIG. 2, the user terminal 200-2 accesses the actor casting platform 100 and inputs evaluation information about a pre-registered actor (S10). Evaluation information about an actor may be an image or characteristic of the actor's appearance.

Although only the process of inputting evaluation information for a previously registered actor is shown in FIG. 2, the user terminal 200-2 can register a new actor in the actor pool and input evaluation information for the actor. The newly registered actor then undergoes image evaluation by other users through step S10.

When actor evaluation information is collected from several user terminals 200-2, the actor casting platform 100 analyzes the collected actor evaluation information and generates image keywords for each actor (S12).

The actor casting platform 100 can generate image keywords by analyzing actor evaluation information through an unsupervised learning model. Through this image keyword analysis process, multiple image keywords can be assigned to each actor.

Next, the casting official terminal 200-1 accesses the actor casting platform 100 and inputs setting information for actor casting (S14).

Setting information for actor casting includes gender, age, physical information, character/image information, and role information. Physical information can be height, weight, body type, presence of double eyelids, etc.; character/image information can be liveliness, cuteness, roughness, quietness, shyness, intelligence, softness, charisma, etc.; role information can be doctor, They can be murderers, soldiers, police officers, con artists, kings, generals, gangsters, etc.

When setting information is input from the casting official terminal 200-1, the actor casting platform 100 executes an actor recommendation process (S16).

The actor casting platform 100 analyzes setting information through a previously learned actor recommendation model, outputs scores for each actor registered in the actor pool, and provides casting information consisting of top-ranked actors to the casting official terminal 200-1. (S18).

In Figure 2, it is explained that the actor recommendation process step (S16) is executed after the image keyword process step (S12). However, this is for convenience of explanation, and the image keyword process step (S12) and the actor recommendation process step (S16) are performed. The order is not set and can be changed or combined at any time.

Figure 3 shows the internal configuration of the actor casting platform according to the present invention.

Referring to FIG. 3, the actor casting platform 100 includes a communication interface 102, a control unit 104, an actor database 106, and the like.

The communication interface 102 is a part for performing data communication with computing devices including the casting agent terminal 200-1 and the user terminal 200-2. The communication interface 102 is a communication module such as Ethernet that is connected to the Internet and performs TCP/IP communication.

The control unit 104 is a part that controls and manages the overall operation of the actor casting platform 100. The control unit 104 performs various functions, but according to an embodiment of the present invention, it performs functions such as an actor recommendation process and an actor image evaluation process.

The control unit 104 includes both hardware and software, and functions such as the actor recommendation process and actor image evaluation process described above may be configured as software modules. The actor recommendation process and the actor image evaluation process are executed by the actor recommendation unit 110 and the actor evaluation unit 120, respectively.

The actor database (DB) 106 is a storage that stores the actor pool. Learning about the actor recommendation unit 110 can be performed using the actor pool information stored in the actor database 106. When general users register a new actor, it is stored in the actor database 106 and the actor pool information is updated.

The actor recommendation unit 110 analyzes setting information for actor casting and recommends the actor requested by the casting official terminal 200-1. The actor recommendation unit 110 according to the present invention analyzes the setting information of casting officials through a supervised learning-based model and outputs casting information consisting of actors optimized for the role.

Referring to FIG. 4, the actor recommendation unit 110 is composed of a feature vector conversion module 112 and a neural network 114.

The feature vector conversion module 112 receives settings information for actor casting, such as gender, age, physical information, character/image information, and role information, and generates a feature vector that can be applied to the deep neural network 112. ) is converted to A feature vector is an N-dimensional value in which each piece of information is normalized.

The N-dimensional feature vector is input to the input layer of the neural network 112 and is converted to a plurality of hidden layers using the weight and bias values and activation function determined through the learning process. layer) and then output to the output layer. The value output to the output layer constitutes casting information.

In order to implement the artificial intelligence-based neural network 114 according to the present invention, training data using gender, age, physical information, character/image information, role information, etc. as input data, and casting information as output data. The neural network 114 is learned through machine learning.

The actor evaluation unit 120 collects actor evaluation information from the user terminal 200-2, analyzes it using an unsupervised learning model, and outputs image keywords for each actor. Actor image keywords generated by the actor evaluation unit 120 are stored in the actor database 106 so that information about each actor can be updated.

Figure 5 shows the processing process of the artificial intelligence-based actor casting method according to the present invention.

Referring to Figure 5, the artificial intelligence-based actor casting method according to the present invention largely consists of an image keyword generation step (S100), an actor recommendation model learning step (S200), and a casting information output step (S300). The above steps are performed on an actor casting platform.

In the image keyword creation step (S100), actor evaluation information for already registered actors is collected from the user terminal, analyzed using an unsupervised learning model, and image keywords for each actor are output.

Image keywords for actors serve as objective indicators of the image perceived by the general public based on the actor's appearance, and can be used to learn an actor recommendation model.

That is, in the actor recommendation model learning step (S200), gender, age, physical information, character/image information based on the image keyword, and role information are used as input data, and a cast consisting of at least one actor among the already registered actors is used as input data. An actor recommendation model is trained using the information as output data.

Once the actor recommendation model learning step (S200) is completed, in the subsequent casting information output step (S300), when gender, age, physical information, character/image information, and role information are input as setting information for actor casting from the casting personnel terminal, Setting information is analyzed through the learned actor recommendation model and casting information is output.

The above description is merely an exemplary description of the present invention, and various modifications may be made by those skilled in the art without departing from the technical spirit of the present invention.

Accordingly, the embodiments disclosed in the specification of the present invention do not limit the present invention. The scope of the present invention should be interpreted in accordance with the scope of the patent claims below, and all technologies within the equivalent scope thereof should be interpreted as being included in the scope of the present invention.

100: Actor casting platform 102: Communication interface
104: Control unit 106: Actor database (DB)
110: Actor recommendation unit 112: Feature vector conversion module
114: Neural network 120: Actor evaluation unit
200: computing device

Claims (4)

When gender, age, physical information, character/image information, and role information are input as setting information for actor casting from the casting official terminal, the actor extracts feature vectors from each information, analyzes the feature vectors through a neural network, and outputs casting information. Recommendation department,
An artificial intelligence-based actor casting platform that includes an actor evaluation unit that analyzes actor evaluation information about registered actors collected from user terminals and outputs image keywords. According to paragraph 1,
The neural network of the actor recommendation unit is an artificial artificial intelligence system characterized in that it is machine-learned using learning data that uses gender, age, physical information, character/image information, and role information as input data and casting information consisting of top-ranked actors as output data. Intelligence-based ordering platform. According to paragraph 1,
The actor evaluation unit is an artificial intelligence-based actor casting platform characterized by analyzing actor evaluation information using an unsupervised learning model. In the actor casting method performed on the artificial intelligence-based actor casting platform,
Collecting actor evaluation information for already registered actors from the user terminal, analyzing it using an unsupervised learning model, and outputting image keywords for each actor;
An actor recommendation model is learned using gender, age, physical information, character/image information and role information based on the image keywords as input data, and casting information consisting of at least one actor among the already registered actors as output data. The steps to do so,
A method including the step of receiving gender, age, physical information, character/image information, and role information as setting information for actor casting from a casting person terminal, analyzing them through the learned actor recommendation model, and outputting casting information.

Images