Abstract: An electrical stimulation apparatus includes: a cell cluster container comprising a cell cluster configured to secrete an active material; an indicator container comprising an indicator; a controller configured to apply a stimulation voltage to the cell cluster container based on a sensing signal received from the indicator container; a first entrance connected to the cell cluster container and configured to allow either one or both of the cell cluster to be retracted and a new cell cluster to be injected; and a second entrance connected to the indicator container and configured to allow either one or both of the indicator to be retracted and a new indicator to be injected.

Abstract: An infrared imaging system is provided. The system includes a sensor configured to receive light emitted by a scene and at least a portion of scenic flux to generate image data, a light source configured to provide calibrating light to offset at least a portion of the scene flux, the light source positioned such that an output of the light source is at a pupil of the infrared imaging system, and at least one image processing device. The image processing device is configured to receive the image data generated by the infrared sensor, determine at least one change in the scenic flux as received by the infrared sensor, determine if the at least one change in the scenic flux results in a change in pixel response of the infrared sensor that exceeds a response threshold, and if the change in pixel response exceeds a threshold, generate an updated calibration table.

Abstract: Provided is a light-emitting diode including a first-light emitting cell, a second light-emitting cell, and a third light-emitting cell that are sequentially provided in one direction and configured to emit light of different colors from each other, a first tunnel junction provided between the first light-emitting cell and the second light-emitting cell, the first tunnel junction being configured to electrically connect the first light-emitting cell and the second light-emitting cell and induce lateral current spreading, and a second tunnel junction provided between the second light-emitting cell and the third light-emitting cell, the second tunnel junction being configured to electrically connect the second light-emitting cell and the third light-emitting cell and induce lateral current spreading.

Abstract: A method of synchronizing time between a host device and a storage device is provided.

Abstract: Provided is a display device including a driving substrate, a barrier layer disposed on an upper surface of the driving substrate and including a plurality of recesses, a micro-semiconductor light emitting device disposed in each of the plurality of recesses, and a side reflective structure disposed in the barrier layer and provided adjacent to a sidewall of each of the plurality of recesses.

Abstract: A processor-implemented neural network operation method includes: receiving a first activation gradient and a first threshold corresponding to a layer included in a neural network; sparsifying the first activation gradient based on the first threshold; determining a second activation gradient by performing a neural network operation based on the sparsified first activation gradient; determining a second threshold by updating the first threshold based on the second activation gradient; and performing a neural network operation based on the second activation gradient and the second threshold.

Abstract: A photographing method and apparatus are provided. The photographing method includes: identifying a photographing distance and a photographing angle with respect to the external device based on timings of signals transmitted and received through a communication connection; acquiring a scene image based on an output of a camera of the photographing device; identifying a current composition model based on the photographing distance, the photographing angle and the scene image; acquiring a set of photographing parameters that are associated with the current composition model; identifying a focal length based on the photographing distance and the current composition model; displaying a preview image based on the output of the camera, the photographing parameter and the current composition model; outputting the preview image as a photographing result based on a photographing instruction; and updating the set of photographing parameters based on the photographing parameter.

Abstract: Provided is an apparatus for estimating a target component, the apparatus including a temperature controller configured to modulate temperature of an object, a measurer configured to measure a spectrum for each temperature of the object that changes based on the modulation, and a processor configured to obtain effective optical pathlength vectors corresponding to a temperature change based on the spectrum for each temperature of the object, obtain a representative effective optical pathlength based on the obtained effective optical pathlength vectors, and obtain a target component estimation model based on the obtained representative effective optical pathlength.

Abstract: The present disclosure relates to a method for performing an uplink (UL) transmission by user equipment in a wireless communication system. In particular, the method includes the steps of: receiving an UL grant associated with a first Hybrid Automatic Repeat Request (HARQ) process from a gNB; selecting a second HARQ process among the first HARQ process and the second HARQ process, when there is a first medium access control (MAC) protocol data unit (PDU) which is associated with the first HARQ process and is a pending MAC PDU; generating a second MAC PDU associated with the second HARQ process; and performing the uplink transmission of the second MAC PDU with an identifier (ID) of the second HARQ process.

Abstract: Broadly speaking, the present techniques generally relate to machine learning models comprising neural network layers, in which the quantisation level of each layer of the model can be independently selected at run-time. In particular, the present application relates to a computer-implemented method for analysing input data on a device using a trained machine learning, ML, model, comprising independently selecting a quantisation level for each of a plurality of network layers of the model at runtime. The present application also relates to a computer-implemented method of training a machine learning model so that the quantisation level of each of the plurality of network layers is independently selectable at runtime. A single trained model with a single set of weights can therefore be deployed, with the quantisation of each layer selected at runtime to suit the capabilities of the device and available resource.

Abstract: An optical sensing device includes: an optical sensor provided on a substrate and configured to radiate light to the outside of a housing through an optical window, and to receive the light that returns to the housing through the optical window; and a heat conductive member connected to the optical sensor and the housing and configured to conduct heat from the optical sensor to the housing.

Abstract: A method for multi-connectivity between a plurality of base stations and user equipment includes estimating, at the user equipment, a first round trip time (RTT) taken in transmitting first data to a first base station, estimating, at the user equipment, a second RTT taken in transmitting second data to a second base station, and determining a size of the first data which is to be transmitted to the first base station, based on the first RTT and the second RTT.

Abstract: Provided are semiconductor devices having a three-dimensional stacked structure and methods of manufacturing the same. A semiconductor device includes a plurality of channel structures on a substrate and arranged in a three-dimensional array; a plurality of gate electrodes extending in a direction parallel to the substrate; and a plurality of source and drain electrodes extending in a direction perpendicular to the substrate. The gate electrodes are connected to the channel structures arranged in the direction parallel to the substrate, and the source and drain electrodes are connected to the channel structures arranged in the direction perpendicular to the substrate. The channel structures include a channel layer and a ferroelectric layer on the channel layer.

Abstract: According to various embodiments of the disclosure, an audio processing apparatus includes at least one processor configured to execute one or more instructions to obtain a second audio signal down-mixed from at least one first audio signal, obtain information related to error removal for the at least one first audio signal, de-mix the at least one first audio signal from the down-mixed second audio signal, and reconstruct the at least one first audio signal by applying the information related to the error removal for the at least one first audio signal to the at least one first audio signal de-mixed from the second audio signal. The information related to the error removal having been generated using at least one of an original signal power of the at least one first audio signal or a second signal power of the at least one first audio signal after decoding.

Abstract: An image detecting device includes a color image sensor configured to sense visible light and to output color image data based on the sensed visible light; a first infrared lighting source configured to provide first infrared rays to a subject; a second infrared lighting source configured to provide second infrared rays to the subject; a mono image sensor configured to sense a first infrared light or a second infrared light reflected from the subject and output infrared image data; and an image signal processor configured to, measure an illuminance value based on the color image data, measure a distance value of the subject based on a portion of the infrared image data corresponding to the first infrared light, and obtain an identification image of the subject based on the illuminance value, the distance value, and a portion of the infrared image data corresponding to the second infrared light.

Abstract: Disclosed are an apparatus and a method for Internet of Things (IoT) device security. The method includes unifying a port in a first IoT device for communication, receiving, by the first IoT device, a packet from a second IoT device through the port, identifying whether the packet in the first IoT device is in a preset packet form, verifying content of the packet in the first IoT device when the packet is in the preset packet form, and opening the port for providing a service in the first IoT device when the verifying of the packet content is successful.

Abstract: A method and apparatus with neural architecture search are provided. A processor-implemented method includes obtaining target data, sampling a trained first neural network into a plurality of second neural networks, training each of the second neural networks based on a portion of the target data, and selecting a second neural network satisfying a predetermined condition among the trained second neural networks for performing an inference operation.

Abstract: An image processing device, including a memory configured to store one or more instructions; and a processor configured to execute the one or more instructions stored in the memory to: obtain kernel coefficient information corresponding to each pixel of a plurality of pixels included in a first image, using a convolution neural network including one or more convolution layers, generate a spatially variant kernel including a kernel corresponding to the each pixel, based on a gradient kernel set including a plurality of gradient kernels corresponding to one or more gradient characteristics of the plurality of pixels, and the kernel coefficient information, and generate a second image, by applying the kernel included in the spatially variant kernel to a region centered on the each pixel, and filtering the first image.

Abstract: A neural network operation apparatus and method is provided. The neural network operation apparatus includes a memory configured to store data for a neural network operation, and a processor configured to validate the data based on a determination that the neural network operation should be performed on the data, obtain a real memory address to perform the neural network operation based on a result of the validating and a virtual tensor address of the data, and perform the neural network operation based on the real memory address.

Abstract: A processor-implemented method includes: determining albedo data in a canonical space and depth data in the canonical space based on input image data including an object, using one or more neural network-based extraction models; generating deformed albedo data and deformed depth data by applying a target shape deformation value respectively to the albedo data and the depth data; generating resultant shaded data by performing shading based on the deformed depth data and a target illumination value; generating intermediate image data based on the resultant shaded data and the deformed albedo data; and generating reconstructed image data from the intermediate image data and the deformed depth data based on a target pose value.