Abstract: An image processing apparatus according to an embodiment removes the noise included in the three-dimensional input image, determines the lost information in a process of obtaining the three-dimensional input image, or enhances the resolution of the three-dimensional input image, by using the neural network learned in advance. The image processing apparatus slices the three-dimensional input image along a depth, converts a three-dimensional input image into a two-dimensional input image, and inputs the converted two-dimensional input image into a neural network. The image processing apparatus generates a three-dimensional output image of which the quality of the three-dimensional input image is enhanced, based on the output of the neural network.

Abstract: A method for attenuating a leakage signal in an FMCW radar system and a radar system thereof are provided. The method includes concentrating a phase noise of the leakage signal on a stationary point and attenuating the phase noise based on the concentration of the phase noise on the stationary point.

Abstract: A learning method of an incremental learning model includes setting a searching range of mask weights based on either one or both of a distribution of mask weights of a binary mask corresponding to a filter of a pretrained model and a learning rate-related parameter, identifying a targeted mask weight in the searching range of the mask weights, updating the targeted mask weight based on the binary mask and the pretrained model, and updating a portion of the binary mask based on the updated targeted mask weight and a preset reference value.

Abstract: In accordance with one or more embodiments of the present disclosure, a method of starting a fuel reformer including a heating element and a subsequent autothermal reformer includes contacting a first fluid comprising oxygen with the heating element, passing the first fluid into the autothermal reformer to preheat a reformer catalyst within the autothermal reformer to a first temperature, reducing flow of the first fluid into the autothermal reformer, introducing a fuel into the autothermal reformer subsequent to preheating the reformer catalyst to initiate a partial oxidation reaction and generating additional heat, increasing flow of the fuel and first fluid to initiate autothermal reforming, and controlling the temperature of the reformer catalyst by supplying a cooling fluid, the first fluid, and the fuel and adjusting flow of each.

Abstract: In a method of inspecting a structure, a first ultrasonic signal generated from a target structure by a first laser beam is received. The first ultrasonic signal is generated by providing the first laser beam generated from a first excitation unit to the target structure. A second ultrasonic signal generated from the target structure by a second laser beam different from the first laser beam is received. The second ultrasonic signal is generated by providing the second laser beam generated from a second excitation unit to the target structure. A third ultrasonic signal generated from the target structure by the first and second laser beams is received. The third ultrasonic signal is generated by simultaneously providing the first and second laser beams to the target structure. It is determined whether the target structure is damaged based on first, second and third ultrasonic frequency spectra that are obtained by converting the first, second and third ultrasonic signals, respectively.

Abstract: The present invention relates to a method and an apparatus for quantum key distribution according to a continuous-variable quantum key distribution protocol, which distributes the quantum key in a reverse post-processing manner and, after photon sub traction at a receiver (Bob), detects bit information from a received quantum state to calculate and share the quantum key, such that security can be further enhanced and a cryptographic key generation rate can be increased since the cryptographic key is not exposed to an attacker (Eve).

Abstract: An electronic device and an operating method thereof according to various embodiments are intended to detect a three-dimensional (3D) object based on camera module and radar sensor fusion, and may be configured to extract a first region of interest (RoI) from a first feature map extracted from first sensing data obtained through a camera module, extract a second RoI from a second feature map extracted from second sensing data obtained through a radar sensor, generate a third RoI by fusing the first RoI and the second RoI, and detect a 3D object based on the third RoI.

Abstract: Disclosed is a 2D-3D HJ-TFET made of a material, the band gap of which changes according to the thickness, such as black phosphorous or TMDC, in order to extend Moore's law. More particularly, disclosed are the structure of a 2D-3D HJ-TFET and a method for manufacturing the same, wherein the 2D-3D HJ-TFET is made of a material such as black phosphorous or TMDC such that the same consumes less power, has a high switching speed, can operate in a complementary manner so as to replace a conventional CMOS transistor, and can extend Moore's law.

Abstract: A memory system may include: a memory controller; a plurality of ranks; and a rank shared bus configured to couple the memory controller and the plurality of ranks. Each of the plurality of ranks may include: a plurality of banks; a rank bus coupled to the plurality of banks and configured to selectively transmit data to the rank shared bus or an intermediate buffer and selectively receive data from the rank shared bus or the intermediate buffer; and an intermediate buffer configured to be selectively coupled to the rank bus or the rank shared bus, according to a first signal from the memory controller.

Abstract: An authentication apparatus includes one or more processors configured to temporally implement a neural network, used to extract a feature value from hidden nodes, that is connected to input nodes to which an electrocardiogram (ECG) signal is input so as to share a weight set with the input nodes, and to match the ECG signal and the extracted feature value to a user for registration.

Abstract: A memory system may include: a memory controller; a plurality of ranks; and a rank shared bus configured to couple the memory controller and the plurality of ranks. Each of the plurality of ranks may include: a plurality of banks; a rank bus coupled to the plurality of banks and configured to selectively transmit data to the rank shared bus or an intermediate buffer and selectively receive data from the rank shared bus or the intermediate buffer; and an intermediate buffer configured to be selectively coupled to the rank bus or the rank shared bus, according to a first signal from the memory controller.

Abstract: A method and apparatus for predicting an object of interest of a user receives an input image of a visible region of a user and gaze information including a gaze sequence of the user, generates weight filters for a per-frame segmentation image by analyzing a frame of the input image for input characteristics of the per-frame segmentation image and the gaze information, and predicts an object of interest of the user by integrating the weight filters and applies the integrated weight filter to the per-frame segmentation image.

Abstract: The present invention relates to a dielectric barrier discharge (DBD) plasma reactor for the preparation of C2+ hydrocarbons from methane, wherein the DBD plasma reactor comprises macroporous silica, as a dielectric material, and optionally a photocatalyst that is impregnated into the pores of the macroporous silica.

Abstract: Provided is a memory-based reinforcement learning method and system capable of storing optional information in streaming data. A question-answering (QA) method using memory-based reinforcement learning method includes receiving, in an episodic memory reader (EMR), streaming data about an input context that is input from a user; analyzing, in the EMR, the received streaming data and storing preset semantic information used for QA in an external memory; and, in response to an input of a question from the user, determining, in a pretrained QA model, an answer to the input question based on sematic information stored in the external memory.

Abstract: Disclosed are a gas sensor member, a gas sensor using the same, and manufacturing methods thereof, and specifically, a gas sensor member using a one-dimensional porous metal oxide nanotube composite material having a double average pore distribution in which mesopores (0.1 nm to 50 nm) and macropores (50 nm to 300 nm) are simultaneously formed on the surface of a nanotube through decomposition of a spherical polymer sacrificial template and continuous crystallization and diffusion of a metal oxide and a nanoparticle catalyst embedded in an apoferritin is uniformly loaded in the inside and on the outer wall and inner wall of a one-dimensional metal oxide nanotube through a high-temperature heat treatment, a gas sensor using the same, and manufacturing methods thereof are disclosed.

Abstract: Proposed are a deep learning method and apparatus for the automatic upright rectification of VR content. The deep learning method for the automatic upright rectification of VR content according to an embodiment may include inputting a VR image, to a neural network and outputting orientation information of the VR image through a trained neural network.

Abstract: Various embodiments, which relate to an electronic device implanted in tissue of an animal and for delivering stimulation to the neural tissue and operating method thereof, may be configured to generate a control instruction based on a control signal wirelessly received from an external device, and based on the control instruction, to input stimulation through a neural probe formed with flexible material and led out to a predetermined location of the tissue. According to various embodiments, the stimulation includes chemical stimulation by a fluid type of drug, and the neural probe may be formed to flow drug, and may include at least one fluid tube for inputting the chemical stimulation by being opened at one end of the neural probe.

Abstract: A processor-implemented neural network method includes: determining an adaptive parameter and an adaptive mask of a current task to be learned among a plurality of tasks of a neural network; determining a model parameter of the current task based on the adaptive parameter, the adaptive mask, and a shared parameter of the plurality of tasks; and training the model parameter and an adaptive parameter of a previous task with respect to the current task, wherein the adaptive parameter of the previous task and the shared parameter are trained with respect to the previous task.

Abstract: A quantum hybridization negative differential resistance device having negative differential resistance (NDR) under a low voltage condition using a nanowire based on an organic-inorganic hybrid halide perovskite, and a circuit thereof are provided. The quantum hybridization negative differential resistance device includes a channel formed of an organic-inorganic hybrid halide perovskite crystal and electrodes formed of its inorganic framework and is connected to opposite ends of the channel.

Abstract: A processor-implemented neural network method includes: extracting, by a feature extractor of a neural network, a plurality of training feature vectors corresponding to a plurality of training class data of each of a plurality of classes including a first class and a second class; determining, by a feature sample generator of the neural network, an additional feature vector of the second class based on a mean vector and a variation vector of the plurality of training feature vectors of each of the first class and the second class; and training a class vector of the second class included in a classifier of the neural network based on the additional feature vector and the plurality of training feature vectors of the second class.