US 12,469,504 B2
Truncateable predictive coding
Erik Norvell, Upplands Väsby (SE); and Fredrik Jansson, Sundbyberg (SE)
Assigned to TELEFONAKTIEBOLAGET LM ERICSSON (PUBL), Stockholm (SE)
Filed by Telefonaktiebolaget LM Ericsson (publ), Stockholm (SE)
Filed on Apr. 23, 2024, as Appl. No. 18/643,227.
Application 18/643,227 is a continuation of application No. 17/817,251, filed on Aug. 3, 2022, granted, now 11,978,460.
Application 17/817,251 is a continuation of application No. 17/044,732, granted, now 11,417,348, issued on Aug. 16, 2022, previously published as PCT/EP2019/058681, filed on Apr. 5, 2019.
Claims priority of provisional application 62/652,941, filed on Apr. 5, 2018.
Claims priority of provisional application 62/652,949, filed on Apr. 5, 2018.
Claims priority of provisional application 62/653,078, filed on Apr. 5, 2018.
Prior Publication US 2024/0347066 A1, Oct. 17, 2024
This patent is subject to a terminal disclaimer.
Int. Cl. G10L 19/012 (2013.01); G10L 19/00 (2013.01); G10L 19/008 (2013.01); G10L 19/032 (2013.01); G10L 19/04 (2013.01); G10L 19/06 (2013.01); H04W 76/28 (2018.01)
CPC G10L 19/012 (2013.01) [G10L 19/0017 (2013.01); G10L 19/008 (2013.01); G10L 19/032 (2013.01); G10L 19/04 (2013.01); G10L 19/06 (2013.01); H04W 76/28 (2018.02)] 23 Claims
OG exemplary drawing
 
1. An encoder comprising:
memory; and
processing circuitry, wherein the encoder is configured to perform a method comprising:
producing prediction residuals for use by a receiver to produce comfort noise, wherein producing the prediction residuals comprises:
obtaining a coherence vector associated with an audio frame of an audio signal, wherein, the coherence vector comprises a coherence value for each frequency band included in a set of two or more frequency bands; and
for each frequency band in the set of frequency bands:
obtaining an intra-frame prediction of the coherence value for the frequency band;
obtaining an inter-frame prediction of the coherence value for the frequency band;
calculating a weighted prediction of the coherence value for the frequency band using: i) the intra-frame prediction of the coherence value for the frequency band, ii) the inter-frame prediction of the coherence value for the frequency band, and iii) a weighting factor; and
producing a prediction residual for the frequency band using the coherence value for the frequency band and the weighted prediction of the coherence value for the frequency band; and
encoding the prediction residuals using a variable bit rate scheme, thereby producing encoded prediction residuals, and
including the encoded prediction residuals in a bitstream.