Dolby International Ab

Dolby Laboratories Licensing Corporation

Arijit Biswas

Michael Schug

Vinay Melkote

Per Hedelin

Harald Mundt

Regunathan Radhakrishnan

Barbara Resch

Lars Villemoes

Jonas Engdegard

Mark Stuart Vinton

Grant Allen Davidson

Tobias Friedrich

Klaus Peichl

Heiko Purnhagen

Kristofer Kjoerling

Christof Fersch

Janusz Klejsa

Aaron Steven Master

Roy M Fejgin

Cong Zhou

Jia Dai

Marco Fink

Simon Plain

Jonas Engdegård

Rolf Meissner

Danilo Hollosi

Ahmed Mustafa

Harald H Mundt

Described herein is a method of determining parameters for a generative neural network for processing an audio signal, wherein the generative neural network includes an encoder stage mapping to a coded feature space and a decoder stage, each stage including a plurality of convolutional layers with one or more weight coefficients, the method comprising a plurality of cycles with sequential processes of: pruning the weight coefficients of either or both stages based on pruning control information, the pruning control information determining the number of weight coefficients that are pruned for respective convolutional layers; training the pruned generative neural network based on a set of training data; determining a loss for the trained and pruned generative neural network based on a loss function; and determining updated pruning control information based on the determined loss and a target loss. Further described are corresponding apparatus, programs, and computer-readable storage media.

Method and apparatus for determining parameters of a generative neural network

Described herein is a method of waveform decoding, the method including the steps of: (a) receiving, by a waveform decoder, a bitstream including a finite bitrate representation of a source signal; (b) waveform decoding the finite bitrate representation of the source signal to obtain a waveform approximation of the source signal; (c) providing the waveform approximation of the source signal to a generative model that implements a probability density function, to obtain a probability distribution for a reconstructed signal of the source signal; and (d) generating the reconstructed signal of the source signal based on the probability distribution. Described are further a method and system for waveform coding and a method of training a generative model.

Methods and system for waveform coding of audio signals with a generative model by implementing a probability distribution

Described herein is a method of processing an audio signal using a deep-learning-based generator, wherein the method includes the steps of: (a) inputting the audio signal into the generator for processing the audio signal; (b) mapping a time segment of the audio signal to a latent feature space representation, using an encoder stage of the generator; (c) upsampling the latent feature space representation using a decoder stage of the generator, wherein at least one layer of the decoder stage applies sinusoidal activation; and (d) obtaining, as an output from the decoder stage of the generator, a processed audio signal. Described are further a method for training said generator and respective apparatus, systems and computer program products.

Method and apparatus for neural network based processing of audio using sinusoidal activation

Described herein is a method of generating a media bitstream to transmit parameters for updating a neural network implemented in a decoder, wherein the method includes the steps of: (a) determining at least one set of parameters for updating the neural network; (b) encoding the at least one set of parameters and media data to generate the media bitstream; and (c) transmitting the media bitstream to the decoder for updating the neural network with the at least one set of parameters. Described herein are further a method for updating a neural network implemented in a decoder, an apparatus for generating a media bitstream to transmit parameters for updating a neural network implemented in a decoder, an apparatus for updating a neural network implemented in a decoder and computer program products comprising a computer-readable storage medium with instructions adapted to cause the device to carry out said methods when executed by a device having processing capability.

Method and apparatus for updating a neural network

A method of decoding an original speech signal for hybrid adversarial-parametric speech synthesis comprising: (a) receiving quantized original linear prediction coding parameters estimated by applying linear prediction coding analysis filtering to an original speech signal and a quantized compressed representation of a residual of the original speech signal; (b) dequantizing the original linear prediction coding parameters and the compressed representation of the residual; (c) inputting the dequantized compressed representation of the residual into a decoder part of a Generator for applying adversarial mapping from the compressed residual domain to a fake (first) signal domain; (d) outputting, by the decoder part of the Generator, a fake speech signal; (e) applying linear prediction coding analysis filtering to the fake speech signal for obtaining a corresponding fake residual; (f) reconstructing the original speech signal by applying linear prediction coding cross-synthesis filtering to the fake residual and the dequantized original linear prediction coding analysis parameters.

Method, apparatus and system for hybrid speech synthesis

Embodiments are directed to a companding method and system for reducing coding noise in an audio codec. A compression process reduces an original dynamic range of an initial audio signal through a compression process that divides the initial audio signal into a plurality of segments using a defined window shape, calculates a wideband gain in the frequency domain using a non-energy based average of frequency domain samples of the initial audio signal, and applies individual gain values to amplify segments of relatively low intensity and attenuate segments of relatively high intensity. The compressed audio signal is then expanded back to the substantially the original dynamic range that applies inverse gain values to amplify segments of relatively high intensity and attenuating segments of relatively low intensity. A QMF filterbank is used to analyze the initial audio signal to obtain a frequency domain representation.

Companding system and method to reduce quantization noise using advanced spectral extension

Described herein is a method of low-bitrate coding of audio data and generating enhancement metadata for controlling audio enhancement of the low-bitrate coded audio data at a decoder side, including the steps of: (a) core encoding original audio data at a low bitrate to obtain encoded audio data; (b) generating enhancement metadata to be used for controlling a type and/or amount of audio enhancement at the decoder side after core decoding the encoded audio data; and (c) outputting the encoded audio data and the enhancement metadata. Described is further an encoder configured to perform said method. Described is moreover a method for generating enhanced audio data from low-bitrate coded audio data based on enhancement metadata and a decoder configured to perform said method.

Method and apparatus for controlling enhancement of low-bitrate coded audio

Embodiments are directed to a companding method and system for reducing coding noise in an audio codec. A method of processing an audio signal includes the following operations. A system receives an audio signal. The system determines that a first frame of the audio signal includes a sparse transient signal. The system determines that a second frame of the audio signal includes a dense transient signal. The system compresses/expands (compands) the audio signal using a companding rule that applies a first companding exponent to the first frame of the audio signal and applies a second companding exponent to the second frame of the audio signal, each companding exponent being used to derive a respective degree of dynamic range compression and expansion for a corresponding frame. The system then provides the companded audio signal to a downstream device.

Coding dense transient events with companding

The present disclosure relates to methods and apparatus for audio coding. A method of encoding a portion of an audio signal comprises determining whether the portion of the audio signal is likely to contain dense transient events, and if it is determined that the portion of the audio signal is likely to contain dense transient events, quantizing the portion of the audio signal using a quantization 5 mode that applies a substantially constant signal-to-noise ratio over frequency for the portion of the audio signal. The present disclosure further relates to a method of detecting dense transient events in a portion of an audio signal.

Low complexity dense transient events detection and coding

An error-concealing audio decoding method comprises: receiving a packet comprising a set of MDCT coefficients encoding a frame of time-domain samples of an audio signal; identifying the received packet as erroneous; generating estimated MDCT coefficients to replace the set of MDCT coefficients of the erroneous packet, based on corresponding MDCT coefficients associated with a received packet directly preceding the erroneous packet; assigning signs of a first subset of MDCT coefficients of the estimated MDCT coefficients, wherein the first subset comprises such MDCT coefficients that are associated with tonal-like spectral bins, to coincide with signs of corresponding MDCT coefficients of said preceding packet; randomly assigning signs of a second subset of MDCT coefficients of the estimated MDCT coefficients, wherein the second subset comprises MDCT coefficients associated with noise-like spectral bins; replacing the erroneous packet by a concealment packet containing the estimated MDCT coefficients and the signs assigned.

MDCT-domain error concealment

Embodiments are directed to a companding method and system for reducing coding noise in an audio codec. A method of processing an audio signal comprises receiving an audio signal, classifying the audio signal as one of pure sinusoidal, hybrid, or pure transient signal using two defined threshold values, and selectively applying a companding operation by switching between a companding off mode, a companding on mode, and an average companding mode, comprising selecting between the companding on mode and the average companding mode for a classified hybrid signal using a companding rule that uses a temporal sharpness measure in a frequency domain.

Signal-dependent companding system and method to reduce quantization noise

Systems and methods are described for switching between lossy coded time segments and a lossless stream of the same source audio. A decoder may receive lossy coded time segments that include audio encoded using frequency-domain lossy coding. The decoder may also receive a lossless stream, which the decoder plays back, that includes audio from the same source encoded using lossless coding. In response to receiving a determination that network bandwidth is constrained, the decoder may generate an aliasing cancellation component based on a previously-decoded frame of the lossless stream, which may be added to a lossy time segment at a transition frame. The sum of the aliasing cancellation component and the lossy time segment may be normalized using a weight caused by an encoding window. Audio playback of the lossy coded time segments may then be provided, beginning with the aliasing-canceled transition frame.

Growing community of inventors

Nuremberg, Germany

Arijit Biswas