US 2026/0290367 A1Published application

Audio de-reverberation

AssigneeDolby Laboratories Licensing Corporation

InventorsJia Dai, Kai Li

Life of this patent

Estimating the term…

Loading patent…

Abstract

Method and system for generating a set of synthesized AIRs from a real acoustic impulse response, AIR(t), and using the set of synthesized AIRs to train a machine learning model, such that the machine learning model, after training, is configured to generate a de-reverberated audio signal given an input audio signal. The synthesized AIRs are generated by forming an early portion, AIRe(t) and a late portion, AIRl(t) of the real AIR by selecting a random separation time point, s, and a random crossfade duration, d. With the proposed approach, a “soft” separation of the real AIR into an early AIR and a late AIR. Specifically, the early AIR will decay to zero during a transition period d, while the late AIR will gradually increase from zero during the transition period. The sum of the early AIR and late AIR will still be equal to the real AIR.

Claims16

Dependent claims

A method for training a machine learning model, the method comprising:

generating a set of synthesized AIRs from a real acoustic impulse response, AIR(t);
using the set of synthesized AIRs to generate a plurality of training samples, each training sample comprising a non-reverberated audio signal and a reverberated audio signal formed by applying one of the synthesized AIRs to the non-reverberated audio signal;
and
training the machine learning model with the plurality of training samples, such that the machine learning model, after training, is configured to generate a de-reverberated audio signal given an input audio signal;
wherein the set of synthesized AIRs is generated by:
forming an early portion, AIRe(t), of the real AIR corresponding to early reflections of a direct sound, and a late portion, AIRl(t), of the real AIR corresponding to late reflections of a direct sound, and
generating each synthesized AIR by modifying at least one of the early portion and the late portion and recombining the possibly modified early portion and possibly modified late portion;
wherein the early portion and the late portion are formed by:
selecting a random separation time point, s,
selecting a random crossfade duration, d,
defining a transition function of time ƒ(t) which describes a continuous decrease from one to zero as t increases from s to s+d,
defining an early crossfade function as:
f
e
(
t
)
=
{
1
t
<
s
f
⁡
(
t
)
s
≤
t
≤
s
+
d
0
t
>
s
+
d
defining a late crossfade function as:
f
l
(
t
)
=
1
-
f
e
(
t
)
,
and
calculating the early portion, AIRe(t), and the late portion, AIRl(t), as:
AIR
e
(
t
)
=
AIR
⁡
(
t
)
*
f
e
(
t
)
⁢
and
⁢
AIR
l
(
t
)
=
AIR
⁡
(
t
)
*
f
l
(
t
)
,
respectively
.

95 more elements

A computer implemented system for training a machine learning model, the system comprising:

a computer implemented process (500) for generating a set of synthesized AIRs from a real acoustic impulse response, AIR(t);
a computer implemented process (900) for generating a plurality of training samples, each training sample comprising a non-reverberated audio signal and a reverberated audio signal formed by applying one of the synthesized AIRs to the non-reverberated audio signal;
and
a computer implemented training process (300) for training the machine learning model using a training set including a plurality of training samples, each training sample comprising a non-reverberated signal and a reverberated signal formed by applying one of the synthesized AIRs to the non-reverberated audio signal;
wherein the computer implemented process for generating a set of synthesized AIRs includes:
a separation block configured to receive a real acoustic impulse response, AIR(t), and to form an early portion, AIRe(t), of the real AIR corresponding to early reflections of a direct sound, and a late portion, AIRl(t), of the real AIR corresponding to late reflections of a direct sound;
at least one processing block for modifying at least one of the early portion and the late portion;
a combination block for recombining the possibly modified early portion and the possibly modified late portion to form a synthesized AIR;
wherein the separation block is configured to:
select a random separation time point, s,
select a random crossfade duration, d,
define a transition function of time ƒ(t) which describes a continuous decrease from one to zero as t increases from s to s+d,
define an early crossfade function as:
f
e
(
t
)
=
{
1
t
<
s
f
⁡
(
t
)
s
≤
t
≤
s
+
d
0
t
>
s
+
d
define a late crossfade function as:
f
l
(
t
)
=
1
-
f
e
(
t
)
,
and
calculate the early portion, AIRe(t), and the late portion, AIRl(t), as:
AIR
e
(
t
)
=
AIR
⁡
(
t
)
*
f
e
(
t
)
⁢
and
⁢
AIR
l
(
t
)
=
AIR
⁡
(
t
)
*
f
l
(
t
)
,
respectively
.

96 more elements

Dependencies are inferred from the claim text.

Facts

Application: 18/874,309Filed 27 Jun 2023
Priority: WO PCT/CN2022/10298430 Jun 2022
Published: 24 Sep 2026
Term ends: Not estimatedThe source records don’t support an estimate.

Classification13 codes

G10L 21/0208
Speech enhancement, e.g. noise reduction or echo cancellation: noise filtering
G06N 20/00
Machine learning
H04S 7/305
Control circuits for electronic adaptation of the sound field: electronic adaptation of stereophonic audio signals to reverberation of the…
G10L 25/30
Characterised by the analysis technique: using neural networks