公开(公告)号：US10818311B2

公开(公告)日：2020-10-27

申请号：US16632373

申请日：2018-11-14

Applicant: INSTITUTE OF AUTOMATION, CHINESE ACADEMY OF SCIENCES

Inventor： Jiaming Xu ,  Jing Shi ,  Bo Xu

IPC: G10L21/00 ,  G10L21/0272 ,  G06F17/16 ,  G06N3/04 ,  G10L25/30

Abstract: An auditory selection method based on a memory and attention model, including: step S1, encoding an original speech signal into a time-frequency matrix; step S2, encoding and transforming the time-frequency matrix to convert the matrix into a speech vector; step S3, using a long-term memory unit to store a speaker and a speech vector corresponding to the speaker; step S4, obtaining a speech vector corresponding to a target speaker, and separating a target speech from the original speech signal through an attention selection model. A storage device includes a plurality of programs stored in the storage device. The plurality of programs are configured to be loaded by a processor and execute the auditory selection method based on the memory and attention model. A processing unit includes the processor and the storage device.

公开(公告)号：US11487950B2

公开(公告)日：2022-11-01

申请号：US16641256

申请日：2019-04-19

Applicant: INSTITUTE OF AUTOMATION, CHINESE ACADEMY OF SCIENCES

Inventor： Jiaming Xu ,  Yiqun Yao ,  Bo Xu

IPC: G06F40/35 ,  G06K9/62 ,  G06N3/04 ,  G06N3/08

Abstract: The method of the present disclosure includes: obtaining an image to be processed and a question text corresponding to the image; using an optimized dialogue model to encode the image into an image vector and encode the question text into a question vector; generating a state vector based on the image vector and the question vector; decoding the state vector to obtain and output an answer text. A discriminator needs to be introduced in an optimization process of the optimized dialogue model. The dialogue model and the discriminator are alternately optimized until a value of a hybrid loss function of the dialogue model and a value of a loss function of the discriminator do not decrease or fall below a preset value, thereby accomplishing the optimization process.

公开(公告)号：US11978470B2

公开(公告)日：2024-05-07

申请号：US17980473

申请日：2022-11-03

Applicant: INSTITUTE OF AUTOMATION, CHINESE ACADEMY OF SCIENCES

Inventor： Jiaming Xu ,  Jian Cui ,  Bo Xu

IPC: G10L21/0272 ,  G10L17/02 ,  G10L17/04 ,  G10L17/06 ,  G10L21/028 ,  H04S1/00

CPC classification number: G10L21/028 ,  G10L17/02 ,  G10L17/04 ,  G10L17/06 ,  H04S1/007

Abstract: Disclosed are a target speaker separation system, an electronic device and a storage medium. The system includes: first, performing, jointly unified modeling on a plurality of cues based a masked pre-training strategy, to boost the inference capability of a model for missing cues and enhance the representation accuracy of disturbed cues; and second, constructing a hierarchical cue modulation module. A spatial cue is introduced into a primary cue modulation module for directional enhancement of a speech of a speaker; in an intermediate cue modulation module, the speech of the speaker is enhanced on the basis of temporal coherence of a dynamic cue and an auditory signal component; a steady-state cue is introduced into an advanced cue modulation module for selective filtering; and finally, the supervised learning capability of simulation data and the unsupervised learning effect of real mixed data are sufficiently utilized.

公开(公告)号：US20230335148A1

公开(公告)日：2023-10-19

申请号：US18026960

申请日：2021-08-24

Applicant: Huawei Technologies Co., Ltd. ,  INSTITUTE OF AUTOMATION, CHINESE ACADEMY OF SCIENCES

Inventor： Henghui Lu ,  Lei Qin ,  Peng Zhang ,  Jiaming Xu ,  Bo Xu

IPC: G10L21/0208 ,  G06V20/40 ,  G10L21/055

CPC classification number: G10L21/0208 ,  G06V20/46 ,  G10L21/055

Abstract: A speech separation method is provided, and relates to the field of speech. The method includes: obtaining, in a speaking process of a user, audio information including a user speech and video information including a user face; coding the audio information to obtain a mixed acoustic feature; extracting a visual semantic feature of the user from the video information; inputting the mixed acoustic feature and the visual semantic feature into a preset visual speech separation network to obtain an acoustic feature of the user; and decoding the acoustic feature of the user to obtain a speech signal of the user. An electronic device, a chip, and a computer-readable storage medium are provided.

公开(公告)号：US10923136B2

公开(公告)日：2021-02-16

申请号：US16645447

申请日：2019-04-19

Applicant: INSTITUTE OF AUTOMATION, CHINESE ACADEMY OF SCIENCES

Inventor： Jiaming Xu ,  Yating Huang ,  Bo Xu

IPC: G10L25/30 ,  G10L21/0208

Abstract: A speech extraction method based on the supervised learning auditory attention includes: converting an original overlapping speech signal into a two-dimensional time-frequency signal representation by a short-time Fourier transform to obtain a first overlapping speech signal; performing a first sparsification on the first overlapping speech signal, mapping intensity information of a time-frequency unit of the first overlapping speech signal to preset D intensity levels, and performing a second sparsification on the first overlapping speech signal based on information of the preset D intensity levels to obtain a second overlapping speech signal; converting the second overlapping speech signal into a pulse signal by a time coding method; extracting a target pulse from the pulse signal by a trained target pulse extraction network; converting the target pulse into a time-frequency representation of the target speech to obtain the target speech by an inverse short-time Fourier transform.