Abstract: A method performed by an electronic device according to an embodiment, includes determining a target point based on a radar signal, determining a first angle set for the target point by using a first array in a virtual MIMO array and determining a second angle set for the target point by using a second array in the virtual MIMO array, generating angle pairs respectively including a corrected first angle and a corrected second angle, determining a matching score for each of the angle pairs by using a third array corresponding to a portion of the virtual MIMO array, and generating multi-angle information of the target point based on target angle pairs, among the angle pairs, of which corresponding matching scores satisfy a predetermined criterion.

Abstract: A method including clustering second point clouds comprised in second point cloud data of a second timepoint, obtaining output data corresponding to one or more second clusters generated as a result of the clustering by applying the second point clouds to a pretrained neural network, determining whether at least one of the one or more second clusters is a target object based on a reliability factor, tracking first clusters corresponding to the target object in first point cloud data of one or more first timepoints prior to the second timepoint, and training of the pretrained neural network based on the output data corresponding to the first clusters and the first point clouds for each first cluster determined according to a result of the tracking.

Abstract: Disclosed is a sign apparatus for displaying a sign for traffic or a vehicle, where the sign apparatus includes a display configured to display the sign and to permit a radar signal radiated from a source to pass, a filter configured to attenuate a signal of at least one frequency in the radar signal passing through the display and to allow a signal of a remaining frequency to pass, and a reflector configured to reflect the signal passing through the filter, wherein the reflected signal includes encoded identification information of the sign.

Abstract: A processor-implemented method including generating an angular spectrum corresponding to a target point, based on a radar signal received through an antenna array of a radar sensor, identifying a target angle of a peak signal corresponding to a maximum peak responsive to a maximum peak of the angular spectrum satisfying a defined condition, updating the angular spectrum by subtracting a target angular spectrum generated based on the peak signal from the angular spectrum, and generating multi-angle information of the target point, the multi-angle information including one or more target angles, the generating the multi-angle information including iteratively performing additional identifying of the target angle and additional updating of the angular spectrum until an iteratively derived maximum peak of an iteratively updated angular spectrum does not satisfy the defined condition.

Abstract: A method and device for object recognition for information collected from a sensor of a moving object are disclosed. The method may include identifying a target with a confidence level of recognition accuracy less than a threshold, based on first sensing information collected from a moving object, obtaining second sensing information on the target from an interworking moving object, synchronizing the first sensing information and the second sensing information based on time information included in the first sensing information and time information included in the second sensing information, and training a neural network for estimating recognition information of the target from the first synchronized sensing information by using information on the target included in the second synchronized sensing information as ground truth data.

Abstract: An operating method of a radar sensor includes: performing object classification on a detected target; comparing a confidence score of a result of the object classification with a threshold value; determining whether high-resolution information on the detected target is to be generated based on a result of the comparison; in response to a determination that the high-resolution information is to be generated, selecting one or more of a plurality of analog-to-digital converters (ADCs) of the radar sensor, and determining control information for changing a setting of the selected one or more ADCs; changing the setting of the selected one or more ADCs based on the determined control information; generating first intermediate frequency signals based on respective first reception signals and a frequency modulated signal, the first reception signals being signals received through reception antennas after a first transmission signal output by the radar sensor is reflected by the detected target, and the first transmi

Abstract: A method, apparatus, and system with multi-modality sensing are provided. A multi-modality sensor includes a radio detection and ranging (radar) sensor including a radio-frequency integrated circuit (RFIC), and an image sensor, including a sensor array, stacked on a portion of the radar sensor, with at least a circuitry portion of the image.

Abstract: A feature extraction method is provided. The feature extraction method includes applying a first feature extracted from multi-channel input data to a bottleneck-based block included in a first type of path of a neural network and obtaining a second feature including a reduced parameter compared to the first feature, upsampling a derived feature of the second feature obtained based on a layer included in a second type of path of the neural network to correspond to a size of a derived feature of the first feature, and obtaining an intermediate feature applied to a head for a task of the neural network, based on the upsampled derived feature of the second feature and the derived feature of the first feature.

Abstract: A method of determining an elevation angle of a target object includes determining a first elevation angle with respect to the target object based on a first component of an initial radar signal received through an array antenna of a radar sensor, determining an angular window based on the first elevation angle, generating a second direction of arrival (DOA) estimation spectrum by applying the angular window to a first DOA estimation spectrum generated based on a second component of the initial radar signal, and determining a second elevation angle with respect to the target object based on the second DOA estimation spectrum.

Abstract: A radar signal processing method includes: extracting a first chirp sequence signal of a first carrier frequency and a second chirp sequence signal of a second carrier frequency from a radar signal received through an array antenna in a radar sensor; generating a first range-Doppler map by performing frequency conversion on the first chirp sequence signal; detecting a first target cell corresponding to a first target in the first range-Doppler map; determining a first ambiguous Doppler velocity of the first target based on first frequency information of the first target cell; determining a first range of an unambiguously measurable Doppler velocity through the first chirp sequence signal; estimating second ambiguous Doppler velocities based on the first ambiguous Doppler velocity and the first range; and determining a Doppler velocity of the first target by performing partial frequency conversion on the second chirp sequence signal based on the second ambiguous Doppler velocities.

Abstract: A method of measuring a velocity vector of a target object is provided. The method includes: obtaining pieces of radar data obtained by detecting the target object; based on the pieces of radar data, extracting velocity vectors for respective points that correspond to the detected target object; and calculating the velocity vector of the target object based on the velocity vectors.

Abstract: A processor-implemented method includes receiving at least one information comprising either one or both of control information of a vehicle and environment information of the vehicle; determining a driving situation of the vehicle, based on the at least one information; changing control parameters for changing a waveform of radar sensors of a sensing system, based on the determined driving situation of the vehicle, wherein the control parameters comprise an operation mode comprising a chirp mode of the radar sensors and comprise sensor parameters of the radar sensors; and transmitting the changed control parameters to the sensing system in real time.

Abstract: Disclosed is a liveness test method and liveness test apparatus. The liveness test method includes determining a presence of a subject using a radar sensor, performing a first liveness test on the subject based on radar data obtained by the radar sensor, in response to the subject being present, acquiring image data of the subject using an image sensor, in response to a result of the first liveness test satisfying a first condition, and performing a second liveness test on the subject based on the image data.

Abstract: An antenna array apparatus includes: first antenna elements arranged to form a first radiation pattern with a recessed portion at the center of the first radiation pattern; and second antenna elements arranged to form a second radiation pattern with a convex portion at the center of the second radiation pattern.

Abstract: A processor-implemented method of performing an operation using a complex-valued attention network includes: extracting a complex-valued attention weight from complex-valued input data; and determining complex-valued attention data by applying the extracted complex-valued attention weight to the complex-valued input data.

Abstract: A processor-implemented method of performing an operation using a complex-valued attention network includes: extracting a complex-valued attention weight from complex-valued input data; and determining complex-valued attention data by applying the extracted complex-valued attention weight to the complex-valued input data.

Abstract: A processor implement method may include determining a target azimuth angle of the target object based on an initial radar signal received through an array antenna of a radar sensor, generating a correction radar signal by correcting a phase of the initial radar signal based on the target azimuth angle, and determining an elevation angle of the target object based on the correction radar signal.

Abstract: Disclosed is a sign apparatus for displaying a sign for traffic or a vehicle, where the sign apparatus includes a display configured to display the sign and to permit a radar signal radiated from a source to pass, a filter configured to attenuate a signal of at least one frequency in the radar signal passing through the display and to allow a signal of a remaining frequency to pass, and a reflector configured to reflect the signal passing through the filter, wherein the reflected signal includes encoded identification information of the sign.

Abstract: A method including clustering second point clouds comprised in second point cloud data of a second timepoint, obtaining output data corresponding to one or more second clusters generated as a result of the clustering by applying the second point clouds to a pretrained neural network, determining whether at least one of the one or more second clusters is a target object based on a reliability factor, tracking first clusters corresponding to the target object in first point cloud data of one or more first timepoints prior to the second timepoint, and training of the pretrained neural network based on the output data corresponding to the first clusters and the first point clouds for each first cluster determined according to a result of the tracking.

Abstract: A control apparatus includes a Wi-Fi communication module, a processor, a controller that controls a component of a vehicle, and a memory, wherein the memory stores at least one piece of position information corresponding to a position in the vehicle, and at least one registered pattern of a gesture, face, or shape, the Wi-Fi communication module transmits a transmission signal to the position corresponding to the at least one piece of position information by a Wi-Fi communication method at a preset frequency and receives a reflection signal reflected from an object in response to the transmission signal, the processor acquires an input pattern of a gesture, face, or shape of the object by analyzing the reflection signal, and retrieves a registered pattern matching the input pattern from the at least one registered pattern, and the controller controls the component according to an instruction corresponding to the retrieved registered pattern.