Abstract: Disclosed is a computing device, which includes a processor, a camera that captures RGB image data, a memory that stores image data including the RGB image data, and a pupil detection device that detects a pupil from the RGB image data stored in the memory in response to a request of the processor. The pupil detection device includes an image converter that receives the RGB image data, detects first eye area data including eye area information from the received RGB image data, and converts the first eye area data into second eye area data having IR (InfraRed) image characteristics by using a deep neural network, a pupil candidate detector that detects pupil candidate data from the second eye area data, and a pupil boundary detector that detects pupil boundary data from the pupil candidate data.

Abstract: Provided is a method of searching for optimal neural network architecture based on channel concatenation. The method includes adjusting spatial size information of an input feature map candidate group so that the spatial size information of the input feature map candidate group corresponds to spatial size information of an output feature map, performing a channel-based concatenation operation on the input feature map candidate group, and additionally extending an output feature map that is the results of the channel-concatenated operation to the input feature map candidate group.

Abstract: Disclosed are a training method for a learning model for recognizing an acoustic signal, a method of recognizing an acoustic signal using the learning model, and devices for performing the methods. The method of recognizing an acoustic signal using a learning model includes identifying an acoustic signal including an acoustic event or acoustic scene, determining an acoustic feature of the acoustic signal, dividing the determined acoustic feature for each of a plurality of frequency band intervals, and determining the acoustic event or acoustic scene included in the acoustic signal by inputting the divided acoustic features to a trained learning model.

Abstract: An apparatus which analyses a depth of a holographic image is provided. The apparatus includes an acquisition unit that acquires a hologram, a restoration unit that restores a three-dimensional holographic image by irradiating the hologram with a light source, an image sensing unit that senses a depth information image of the restored holographic image, and an analysis display unit that analyzes a depth quality of the holographic image, based on the sensed depth information image, and the image sensing unit uses a lensless type of photosensor.

Abstract: The present invention relates to a method and an apparatus for binaural rendering an audio signal using variable order filtering in frequency domain. To this end, provided are a method for processing an audio signal including: receiving an input audio signal; receiving a set of truncated subband filter coefficients for filtering each subband signal of the input audio signal, the set of truncated subband filter coefficients being constituted by one or more FFT filter coefficients generated by performing FFT by a predetermined block size; generating at least one subframe for each subband; generating at least one filtered subframe for each subband; performing inverse FFT on the filtered subframe for each subband; and generating a filtered subband signal by overlap-adding the transformed subframe for each subband and an apparatus for processing an audio signal using the same.

Abstract: Provided are a pulse-matched filter-based packet detection apparatus and method. The packet detection apparatus includes a photoelectric converter for converting an optical wireless communication signal into an electrical signal, an analog-to-digital converter (ADC) for converting the electrical signal into a digital signal, a pulse-matched filter for outputting a first correlation representing a correlation between a pulse obtained by oversampling a modulated pulse and the digital signal, a correlator for outputting a second correlation representing a correlation between the first correlation and a preamble code, a packet detection signal generator for generating a packet detection signal by comparing the second correlation and a packet detection threshold value, and a demodulator for demodulating the digital signal based on the packet detection signal.

Abstract: There is provided an video encoding/decoding method and apparatus. The video decoding method comprises acquiring a bitstream including a predetermined context element, performing at least one of a context model determination, a probability update, and a probability interval determination on the predetermined syntax element, and arithmetically decoding the predetermined syntax element on the basis of a result of the performance.

Abstract: Disclosed herein are an apparatus and method for monitoring the duty cycle of a memory clock signal. The apparatus for monitoring a duty cycle of a memory clock signal includes a clock frequency converter configured to generate a second monitoring target clock signal by decreasing a frequency of a first monitoring target clock signal while maintaining a duty cycle of the first monitoring target clock signal, and a pulse counter configured to measure a pulse width of the second monitoring target clock signal using a reference clock signal.

Abstract: A method of moving a magic qubit in logical qubit arrangement architecture based on a checkerboard may include generating a plurality of distilled magic qubits necessary for a predetermined logic operation through a magic state distiller, storing the plurality of distilled magic qubits in a magic qubit repository, primarily moving the stored magic qubit to an area adjacent to a logical qubit block corresponding to a destination through an external bus and an internal bus, and secondarily moving the magic qubit from the area adjacent to the logical qubit block to the destination within the logical qubit block through a temporary bus.

Abstract: Disclosed is a method for realistic visualization of a digital human, the method including: setting a specific action of the digital human; determining a scene including the specific action of the digital human and rendering the determined scene to generate a first rendered video; capturing images constituting the first rendered video for each frame to obtain frame data; inputting each piece of the frame data of the first rendered video to two or more realistic visualization modules to obtain frame data of a second rendered video; and combining the frame data of the second rendered video to generate a realistically visualized scene.

Abstract: A semiconductor package is provided. The semiconductor package includes a semiconductor chip including an integrated circuit (IC) pad configured to output a wireless frequency signal, a lead frame including a package pad configured to receive the wireless frequency signal, a first wire of which one end is connected to the IC pad, a second wire of which one end is connected to the package pad, and a circuit element connected to the first wire and the second wire in parallel or in series. A radio frequency (RF) output impedance of the package pad or an RF output impedance of the IC pad is controlled by a type of the circuit element or a connection relationship between the circuit element, the first wire, and the second wire.

Abstract: Disclosed is an artificial intelligence apparatus for detecting a target gas, which includes a mixed gas measurement unit that measures a mixed gas collected in a plurality of domains through a sensor array to generate sensing data including heterogeneous domain measurement data measured from the mixed gas collected in a domain different from the target gas and target domain measurement data measured from the mixed gas collected from the same domain as the target gas, a heterogeneous intelligence model deep learning unit that receives the heterogeneous domain measurement data to train a heterogeneous intelligence model, a target intelligence model deep learning unit that receives the heterogeneous intelligence model and the target domain measurement data to train a target intelligence model, and a target gas detection unit that determines whether an environmental gas includes the target gas using the target intelligence model.

Abstract: A memory access method and device are provided. A memory access method may include: identifying, when an access to a page of a remote memory occurs, a type of the access; allocating a sparse buffer when the access is a sparse write; storing data for the sparse write in the sparse buffer; storing an address for the sparse write as a key and the sparse buffer as a value in a buffer table; and updating an instruction pointer to point to a next instruction.

Abstract: An encoding method and decoding method, and a device implementing the same are provided. The encoding method obtains an intra prediction mode related to a current block from a video signal, obtains prediction samples of the current block by performing intra prediction on the current block based on the intra prediction mode, scans transform coefficients of the current block based on a scanning type of the current block, obtains dequantized transform coefficients of the current block by dequantizing the transform coefficients, obtains residual samples of the current block by performing an inverse-transform on the dequantized transform coefficients, the inverse-transform being performed in a horizontal direction and a vertical direction, reconstructs the current block using the residual samples and the prediction samples, and applies a deblocking filter on the reconstructed current block.

Abstract: A method and a device for encoding/decoding images are disclosed. The method for encoding images comprises the steps of: deriving a scan type of a residual signal for a current block according to whether or not the current block is a transform skip block; and applying the scan type to the residual signal for the current block, wherein the transform skip block is a block to which transform for the current block is not applied and is specified on the basis of information indicating whether or not transform for the current block is to be applied.

Abstract: An apparatus for transmitting a broadcast signal according to an embodiment of the present invention includes a bootstrap generation unit for generating a bootstrap for signaling a system bandwidth field, corresponding to a system bandwidth for a post-bootstrap, and a baseband sampling rate coefficient; a preamble generation unit for generating a preamble located immediately following the bootstrap in a broadcast signal frame; and a payload generation unit for generating one or more subframes located immediately following the preamble in the broadcast signal frame. Here, the preamble may include L1 signaling data, and the L1 signaling data may include a reduced carrier coefficient.

Abstract: A method of user equipment may comprise: determining whether an error condition for a first beam communicating with a base station is met; when it is determined that the error condition for the first beam is met, determining whether a first physical uplink control channel (PUCCH) allocated to the user equipment is present within a predetermined first period of time; when it is determined that the first PUCCH is not present, transmitting a sounding reference signal (SRS) transmission approval request and an identifier of the user equipment to the base station through a second PUCCH pre-configured to be shared by all pieces of user equipment; and when an SRS transmission approval is received from the base station, transmitting an SRS to the base station.

Abstract: Disclosed herein is a method for storing blockchain transaction data. The method includes selecting transaction blocks corresponding to an encoding group, generating at least one encoding chunk corresponding to each of participating nodes by performing fountain encoding on the transaction blocks, and storing the at least one encoding chunk corresponding to one of the participating nodes.

Abstract: A method of a base station may comprise: detecting a beam failure for a terminal; in response to detecting the beam failure, transmitting a beam recovery early indicator (BREI) to the terminal based on a beam sweeping scheme; receiving a BREI response from the terminal; and identifying a preferred beam of the terminal based on reception of the BREI response.

Abstract: A method of a base station may comprise: receiving a signal including a first message of a first terminal and a first message of a second terminal in one RO; determining the first terminal as a near terminal based on preconfigured criteria; determining the second terminal as a far terminal based on the preconfigured criteria; transmitting, to the first terminal and the second terminal, one or more DCIs including a NOMA indicator indicating that second messages, which are responses to the first messages of the first terminal and the second terminal, are to be respectively transmitted in a NOMA scheme; and transmitting the second messages to the first terminal and the second terminal based on the one or more DCIs.