摘要:
A method and apparatus for eliminating cross-channel interference, and a multi-channel source separation method and a multi-channel source separation apparatus using the same. The multi-channel source signal separation apparatus includes: a source separation unit separating multi-channel source signals from a mixture including the multi-channel source signals; and a post-processing unit eliminating cross-channel interference from an arbitrary channel output of the separated multi-channel source signals by using an interference elimination coefficient determined based on a degree of interference between the arbitrary channel output and a different channel output of the separated multi-channel source signals.
摘要:
Cross-channel interference is eliminated and multi-channel sources are separated by estimating a source absence probability for a current frame of a first channel output, and determining an interference elimination coefficient for matching a secondary signal of the first channel output with a primary signal of a second channel output by using the source absence probability, generating an interference signal by multiplying the second channel output by an over-subtraction factor and the interference elimination coefficient, wherein a partial differentiation is performed for a v-norm value of a spectral amplitude difference, between the first channel output and the second channel output multiplied by the interference elimination coefficient and a result of multiplication of the source absence probability, by using the interference elimination coefficient to determine an update amount of the interference elimination coefficient for a next frame.
摘要:
An apparatus for tracking and identifying objects includes an audio likelihood module which determines corresponding audio likelihoods for each of a plurality of sounds received from corresponding different directions, each audio likelihood indicating a likelihood a sound is an object to be tracked; a video likelihood module which receives a video and determines video likelihoods for each of a plurality of images disposed in corresponding different directions in the video, each video likelihood indicating a likelihood that the image is an object to be tracked; and an identification and tracking module which determines correspondences between the audio likelihoods and the video likelihoods, if a correspondence is determined to exist between one of the audio likelihoods and one of the video likelihoods, identifies and tracks a corresponding one of the objects using each determined pair of audio and video likelihoods.
摘要:
A person tracking method and apparatus using a robot. The person tracking method includes: detecting a person in a first window of a current input image using a skin color of the person; and setting a plurality of second windows in a next input image, correlating the first window and the second windows and tracking the detected person in the next input image using the correlated results.
摘要:
An apparatus and method of controlling a mobile body that travels around a sound source which generates a sound. This apparatus includes a traveling information producer for producing traveling information, which is information about traveling of the mobile body; a direction estimator for estimating a direction in which the mobile body is located with respect to the sound source; and a position determiner for determining a position of the mobile body using the traveling information and the estimated direction of the mobile body.
摘要:
A noise elimination method and apparatus. The method eliminates noise from an input signal containing a voice signal mixed with a noise signal. The method includes detecting a noise section, in which the noise signal is present, from the input signal; obtaining a weight to be used for the input signal from signals of the noise section; and filtering the input signal using the obtained weight. The method and apparatus enable a mobile robot to eliminate noise in real time and effectively detect and recognize voice.
摘要:
A person tracking method and apparatus using a robot. The person tracking method includes: detecting a person in a first window of a current input image using a skin color of the person; and setting a plurality of second windows in a next input image, correlating the first window and the second windows and tracking the detected person in the next input image using the correlated results.
摘要:
An apparatus and method of controlling a mobile body that travels around a sound source which generates a sound. This apparatus includes a traveling information producer for producing traveling information, which is information about traveling of the mobile body; a direction estimator for estimating a direction in which the mobile body is located with respect to the sound source; and a position determiner for determining a position of the mobile body using the traveling information and the estimated direction of the mobile body.
摘要:
A method and apparatus for robust speaker localization and a camera control system employing the same are provided. The apparatus for speaker localization includes: a difference spectrum obtaining section which obtains a difference spectrum of a first pseudo-power spectrum for a speech section and a second pseudo-power spectrum for a non-speech section detected in a voice signal output from a microphone array; and a speaker direction estimation section which detects a peak value in any one of the difference spectrum and the first pseudo-power spectrum, and estimates the direction of a speaker based on the direction angle corresponding to the detected peak value.
摘要:
An adaptive beamforming apparatus and method includes a fixed beamformer that compensates for time delays of M noise-containing speech signals input via a microphone array having M microphones (M is an integer greater than or equal to 2), and generates a sum signal of the M compensated noise-containing speech signals; and a multi-channel signal separator that extracts pure noise components from the M compensated noise-containing speech signals using M adaptive blocking filters that are connected to M adaptive canceling filters in a feedback structure and extracts pure speech components from the added signal using the M adaptive canceling filters that are connected to the M adaptive blocking filters in the feedback structure.