Abstract: A camera module includes a housing, a reflective member positioned in the housing and changing a direction of light to a direction of an optical axis, a carrier carrying the reflective member and rotatable about a first axis with respect to the housing, and a first ball group disposed between the housing and the carrier, wherein the first ball group includes a main ball member providing the first axis of the carrier, and an auxiliary ball member disposed away from the first axis, and one or both of the housing and the carrier partially accommodates the auxiliary ball member, and includes an auxiliary guide groove extended in a circumferential direction of the first axis.

Abstract: A protective film is provided. The protective film includes a release film, a base film which is disposed on the release film and comprises a protective part and a first pull tab part protruding from a first side surface of the protective part, and a first dummy film which is disposed on the release film, does not overlap the base film in a plan view, and comprises a part partially surrounding the first pull tab part, where edges of the base film and edges of the first dummy film are disposed inside edges of the release film in the plan view.

Abstract: Disclosed herein are a neural network model deployment method and apparatus for providing a deep learning service. The neural network model deployment method may include providing a specification wizard to a user, searching for and training a neural network based on a user requirement specification that is input through the specification wizard, generating a neural network template code based on the user requirement specification and the trained neural network, converting the trained neural network into a deployment neural network that is usable in a target device based on the user requirement specification, and deploying the deployment neural network to the target device.

Abstract: A display device includes: a display panel including a display area including pixels and a non-display area including a dummy pixel; a scan driver which supplies a scan signal to the display panel; a data driver which supplies a data signal to the display panel; and a timing controller which supplies a first control signal for controlling the scan driver and a second control signal for controlling the data driver. The dummy pixel is connected to a bad pixel among the pixels in the display area through a repair line, and a connection of the dummy pixel to the repair line is cut off in an initialization phase in which a voltage of an initialization power source is supplied.

Abstract: Disclosed herein are an apparatus and method for generating a neural network executable image. The apparatus includes one or more processors and executable memory for storing at least one program executed by the one or more processors. The at least one program receives user requirements including a default neural network model and training result data for generating a neural network executable image required by a user, checks whether the default neural network model included in the user requirements is capable of being supported in a target system in which the neural network executable image is to be installed, converts the default neural network model into a neural network model executable in the target system, converts the training result data by reconfiguring the data format set of the training result data, and generates a neural network executable image by combining the converted neural network model and the converted training result data.

Abstract: Disclosed herein are an apparatus and method for developing a neural network application. The apparatus includes one or more processors and executable memory for storing at least one program executed by the one or more processors. The at least one program receives a target specification and an application specification including user requirements, searches for a neural network model corresponding to the target specification and the application specification in a database, builds an inference engine for performing a neural network operation used by the neural network model, and generates a target image for executing the neural network model to be suitable for a target device using the inference engine.

Abstract: Disclosed herein are a federated learning method and apparatus. The federated learning method includes receiving a feature vector extracted from a client side and label data corresponding to the feature vector, outputting a feature vector with phase information preserved therein by applying the feature vector as input of a Self-Organizing Feature Map (SOFM), and training a neural network model by applying both the feature vector with the phase information preserved therein and the label data as input of a neural network model.

Abstract: An embodiment relates to an artificial intelligence inference apparatus and method. The embodiment provides an artificial intelligence inference method, and may include converting an application based on a previously learned neural network into executable code in a high-level language independent of a learning framework, separating the executable code into General-Purpose Language (GPL) code and Domain-Specific Language (DSL) code depending on whether an acceleration operation is required, and generating target code optimized for hardware from the separated GPL code and DSL code.

Abstract: Disclosed herein are an apparatus and method for converting a neural network. The method includes separating neural network data of a source framework to form a tree structure by analyzing the same, converting the neural network data in a tree structure to a neural network optimized for a target framework, classifying training data based on the result of analysis of the neural network data of the source framework, converting the classified training data to the training data structure of the target framework, and creating a neural network and training data of the target framework by combining the converted neural network and the converted training data.

Abstract: A cost-cognitive base station apparatus, a cost-cognitive user terminal, and a cost-cognitive network method, in which by recognizing costs based on network congestion, a user may select and receive appropriate network services, thereby reducing telecommunication expenses and inducing the user to a less congested network, which leads to improved network management efficiency.

Abstract: The present invention relates to smart glasses, and the smart glasses include: a frame mounted with lens; a pair of temples extended in one direction from both side ends of the frame so as to face each other; a rear camera mounted at an end of the temple and configured to photograph a rear image of a wearer; a display unit installed at one side of the lens and configured to display the rear image photographed by the rear camera; and a control device configured to control an operation of each of the constituent elements, thereby enabling a user to check a situation in the rear view without turning his/her head.

Abstract: An apparatus and a method for managing instant connection based on a wireless local area network are provided. The method includes: creating, by an accessed terminal, connection information required for setting connection therewith; creating, by the accessed terminal, a management frame by using the connection information; transmitting, by the accessed terminal, the created management frame to a peripheral terminal; receiving, by the peripheral terminal, the management frame and extracting connection information from the received management frame; and setting, by the peripheral terminal, connection with the accessed terminal by accessing the accessed terminal by using the extracted connection information to rapidly access a desired terminal without performing a device discovery process.

Abstract: Disclosed herein are a location measurement method and apparatus. The apparatus includes a first grading unit, a first presumed line calculation unit, a second grading unit, a second presumed lined calculating unit, a presumed location calculation unit, and a final location calculation unit. The first grading unit determines the grade of a first RSSI. The first presumed line calculation unit calculates the range of the object from a first node based on the grade of the first RSSI. The second grading unit determines the grade of a second RSSI. The second presumed line calculating unit calculates the range of the object from a second node based on the grade of the second RSSI. The presumed location calculation unit calculates two presumed locations. The final location calculation unit determines one of the two presumed locations to be the final location of the object.