Abstract: A method of providing a video stream based on machine learning in an electronic device according to various example embodiments may include receiving a source video stream which is streamed from a first device to at least one other device, confirming whether an event is detected on the source video stream using a learning model trained through machine learning on the basis of at least one frame of the source video stream, and determining whether to restrict streaming of the source video stream from the first device on the basis of the event detection. In addition to the method, other example embodiments are possible.

Abstract: A method of providing a video stream based on machine learning in an electronic device according to various example embodiments may include receiving a source video stream which is streamed from a first device to at least one other device, confirming whether an event is detected on the source video stream using a learning model trained through machine learning on the basis of at least one frame of the source video stream, and determining whether to restrict streaming of the source video stream from the first device on the basis of the event detection. In addition to the method, other example embodiments are possible.

Abstract: The operating method of a server for providing a metaverse in which a call between user terminals may be performed may include: generating a first avatar based on profile information of a first user terminal and a second avatar based on profile information of a second user terminal; providing a map space in which the first avatar and the second avatar are able to travel; mediating a call between the first user terminal and the second user terminal based on whether a plurality of avatars including the first avatar and the second avatar have entered a first audio space included in the map space; and completing matching between the first user terminal and the second user terminal when a first sign of attraction to the second avatar is transmitted from the first user terminal and a sign of attraction to the first avatar is transmitted from the second user terminal.

Abstract: Provided is an operation method of a server for relaying a call. The server relays a preliminary call between a first terminal and a second terminal for a preset time, based on matching of a first user of the first terminal and a second user of the second terminal, and after the preset time has elapsed, receives, from the first terminal and the second terminal, a response as to whether to continue the preliminary call. The server may determine whether to proceed with a main call based on the received response.

Abstract: Provided is an operation method for a terminal performing a call. A first terminal may receive a first like signal transmitted from a second terminal to a server, and may transmit, to the server, a second like signal regarding a second user of the second terminal. At this time, the first terminal may receive a signal for establishing a call connection between the first terminal and the second terminal, and may perform a call with the second terminal based on the signal.

Abstract: A mediating method by a mediating device comprises receiving a first list including at least one item from a first terminal and receiving at least one list from at least one terminal different from the first terminal, searching for a list including an item corresponding to at least one item among the at least one item of the first list, among the received at least one list, and transmitting information on the corresponded item, according to a result of the search, to each of the first terminal and a terminal which provides the list including the item corresponding to the at least one item among the at least one terminal.

Abstract: A keyword spotting apparatus, method, and computer-readable recording medium are disclosed. The keyword spotting method using an artificial neural network according to an embodiment of the disclosure may include obtaining an input feature map from an input voice; performing a first convolution operation on the input feature map for each of n different filters having the same channel length as the input feature map, wherein a width of each of the filters is w1 and the width w1 is less than a width of the input feature map; performing a second convolution operation on a result of the first convolution operation for each of different filters having the same channel length as the input feature map; storing a result of the second convolution operation as an output feature map; and extracting a voice keyword by applying the output feature map to a learned machine learning model.

Abstract: Provided is a terminal. The terminal may include an input and output interface configured to receive a first video capturing a first user, a communication interface configured to transmit the first video or a first sound of the first user to a terminal of a counterpart and receive a second video or second sound, which is obtained by the terminal of the counterpart, through a video call session, a display configured to display the first video and the second video, and a processor configured to detect a specific event during the video call session, determine whether an inappropriate element is included in the first video or the first sound when the specific event is detected, and stop transmitting the first video or the first sound according to a result of the determination.

Abstract: One or more embodiment relates to a video communication device, video communication method, and a video communication mediating method, which are capable of preventing a user in a video communication from being exposed to an inappropriate image or sound from the other party. One or more embodiment may provide a video communication method comprising establishing a first video communication session with a first terminal, consecutively receiving images or sounds from the first terminal through the first video communication session, examining at least a portion of images or sounds received in a period of recent specific time among the received images or sounds, and blocking an output of at least a portion of the received images or sounds or reporting a user of the first terminal as an abusive user to a first server according to the result of the examining.

Abstract: A video call mediation method of a system may include receiving, by a server, a mediation request, from a plurality of mobiles, mediating, by the server, a first mobile and a second mobile, of the plurality of mobiles, establishing, by the first mobile and the second mobile, a video call session, receiving, by the first mobile, a video, from the second mobile, through the video call session, detecting, by the first mobile, a certain input, reporting, by the first mobile, a video received in the server, in response to the certain input, ending, by the first mobile, the video call session with the second mobile, and establishing, by the first mobile, a video call session with a third mobile, and verifying, by the server, the reporting, and rejecting, by the server, additional mediation request of the second mobile.

Abstract: Provided is an apparatus for evaluating and improving responses, and a method and a computer readable recording medium thereof. The apparatus for evaluating responses according to the present disclosure obtains cluster classifying information for training responses, and based on distribution of clusters to which test responses output from the dialogue generation model are classified, evaluate semantic diversity of the responses output from the dialogue generation model.

Abstract: Provided is an apparatus for evaluating and improving responses, and a method and a computer readable recording medium thereof. The apparatus for evaluating responses according to the present disclosure obtains cluster classifying information for training responses, and based on distribution of clusters to which test responses output from the dialogue generation model are classified, evaluate semantic diversity of the responses output from the dialogue generation model.

Abstract: A report evaluation method of a report evaluation device may include receiving a report from at least one client terminal, determining a category of the received report, identifying a learning model corresponding to the category, evaluating a reliability of the report through the learning model, and generating and outputting information on the reliability.

Abstract: Systems and techniques to generate imitative responses are illustrated. response generation method performed in an electronic apparatus of the present disclosure includes acquiring at least one piece of utterance data, acquiring a first context corresponding to the utterance data from a context candidate set, generating one or more dialogue sets including the first context and the utterance data, receiving a second context from a user, and acquiring a response corresponding to the second context using a language model based on the one or more dialogue sets.

Abstract: Provided is a method for training a model, including generating a plurality of attention maps by inputting training data into a previously trained teacher model, generating a set of attention weights of the teacher model based on the plurality of attention maps, generating a set of attention weights of a student model by inputting the training data into the student model, calculating a value of a first loss function based on the set of attention weights of the teacher model and the set of attention weights of the student model, calculating a value of a second loss function according to an inference of the student model with respect to the training data, and training the student model based on the value of the first loss function and the value of the second loss function.

Abstract: The operating method of a server for providing a metaverse in which a call between user terminals may be performed may include: generating a first avatar based on profile information of a first user terminal and a second avatar based on profile information of a second user terminal; providing a map space in which the first avatar and the second avatar are able to travel; mediating a call between the first user terminal and the second user terminal based on whether a plurality of avatars including the first avatar and the second avatar have entered a first audio space included in the map space; and completing matching between the first user terminal and the second user terminal when a first sign of attraction to the second avatar is transmitted from the first user terminal and a sign of attraction to the first avatar is transmitted from the second user terminal.

Abstract: Disclosed are a terminal, an operating method thereof, and a computer-readable recording medium. The operating method includes establishing a video call session between a first terminal and a second terminal, receiving, by the second terminal, data obtained by the first terminal, sequentially storing, by the second terminal, the data received from the first terminal in a buffer size, performing, by the second terminal, a validity check on the data stored in the buffer, and processing, by the second terminal, the data in response to a result of the validity check.

Abstract: A terminal may include a display that is divided into at least two areas, when a real time broadcasting, where a user of the terminal is a host, starts through a broadcasting channel, and of which one area of the at least two areas is allocated to the host; an input/output interface that receives a voice of the host; a communication interface that receives one item selected of at least one or more items and a certain text from a terminal of a certain guest, of at least one or more guests who entered the broadcasting channel; and a processor that generates a voice message converted from the certain text into the voice of the host or a voice of the certain guest.