Abstract: A method of matching a voice for each object included in a video, includes: separating a plurality of voices in a video; determining a dissimilarity between the plurality of voices; selecting a partial duration in an entire duration of the video as a matching duration, based on the dissimilarity between the plurality of voices; matching, within the matching duration, the plurality of voices with a plurality of objects in the video respectively, based on mouth movements of the plurality of objects; and matching the plurality of voices with the plurality of objects respectively in the entire duration of the video, based on results of the matching between the plurality of voices and the plurality of objects within the matching duration.

Abstract: An audio signal processing apparatus is configured to: transform a first audio signal includes n channels to generate a first audio data in a frequency domain, generate a frequency feature signal for each channel from the first audio data in the frequency domain, based on a first deep neural network (DNN), generate a second audio signal includes m channels from the first audio signal, based on a second DNN, and generate an output audio signal by encoding the second audio signal and the frequency feature signal. The first audio signal is a high order ambisonic signal includes a zeroth order signal and a plurality of first order signals. The second audio signal includes a mono signal or a stereo signal. m is smaller than n.

Abstract: A video processing apparatus includes a memory storing instructions, and at least one processor configured to execute the instructions to generate a plurality of feature information by analyzing a video signal comprising a plurality of images based on a first DNN, extract a first altitude component and a first planar component corresponding to a movement of an object in a video from the video signal based on a second DNN, extract a second planar component corresponding to a movement of a sound source in audio from a first audio signal based on a third DNN, generate a second altitude component based on the first altitude component, the first planar component, and the second planar component, output a second audio signal comprising the second altitude component based on the feature information, and synchronize the second audio signal with the video signal and output the synchronized second audio signal and video signal.

Abstract: An apparatus for processing audio includes at least one processor configured to obtain a down-mixed audio signal from a bitstream, to obtain down-mixing-related information from the bitstream, to de-mix the down-mixing-related information by using down-mixing-related information, and to reconstruct an audio signal including at least one frame based on the de-mixed audio signal. The down-mixing-related information is information generated in units of frames by using an audio scene type.

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 audio processing apparatus may obtain second audio signals corresponding to channels included in a second channel group from first audio signals corresponding to channels included in a first channel group, downsample at least one third audio signal corresponding to at least one channel identified based on a correlation with the second channel group from among the channels included in the first channel group, by using an artificial intelligence (AI) model, and generate a bitstream including the second audio signals corresponding to the channels included in the second channel group and the downsampled at least one third audio signal. The first channel group includes a channel group of an original audio signal, and the second channel group is constructed by combining at least two channels from among the channels included in the first channel group.