公开(公告)号：US11003413B2

公开(公告)日：2021-05-11

申请号：US14920463

申请日：2015-10-22

Applicant: Google Inc.

Inventor： Christopher Russell LaRosa ,  Sam Kvaalen ,  Thomas Chadwick Walters ,  Richard Francis Lyon ,  Robert Steven Glickstein ,  Rushabh Ashok Doshi ,  Molly Castle Nix ,  Jason Matthew Toff

IPC: G06F17/00 ,  G06F3/16 ,  G11B27/034 ,  G11B27/28 ,  G10L21/0272 ,  G10H1/00 ,  H04H60/04

Abstract: Systems and techniques for removing a sound recording from an audio recording (e.g., an audio recording embedded in a media file) are presented. The system can include an identification component, a first subtraction component and a second subtraction component. The identification component identifies a sound recording in a mixed audio recording. The first subtraction component determines a local linear transformation of the sound recording and subtracts the local linear transformation of the sound recording from the mixed audio recording to generate a new mixed audio recording. The second subtraction component compares one or more segments of the sound recording with one or more corresponding segments of the new mixed audio recording and reduces a power level of the new mixed audio recording based at least in part on correlation of the one or more corresponding segments with the one or more segments.

公开(公告)号：US09195431B2

公开(公告)日：2015-11-24

申请号：US13730076

申请日：2012-12-28

Applicant: Google Inc.

Inventor： Christopher Russell LaRosa ,  Sam Kvaalen ,  Thomas Chadwick Walters ,  Richard Francis Lyon ,  Robert Steven Glickstein ,  Rushabh Ashok Doshi ,  Molly Castle Nix ,  Jason Matthew Toff

IPC: G06F17/00 ,  G06F3/16 ,  G11B27/034 ,  G11B27/28 ,  G10L21/0272 ,  G10H1/00 ,  H04H60/04

CPC classification number: G06F3/16 ,  G10H1/0008 ,  G10H1/0025 ,  G10H2210/021 ,  G10H2210/125 ,  G10L21/0272 ,  G11B27/034 ,  G11B27/28 ,  H04H60/04 ,  Y02D70/1224 ,  Y02D70/1226 ,  Y02D70/1242 ,  Y02D70/1244 ,  Y02D70/1246 ,  Y02D70/1262 ,  Y02D70/1264 ,  Y02D70/142 ,  Y02D70/144 ,  Y02D70/146 ,  Y02D70/168

Abstract: Systems and techniques for removing a sound recording from an audio recording (e.g., an audio recording embedded in a media file) are presented. The system can include an identification component, a first subtraction component and a second subtraction component. The identification component identifies a sound recording in a mixed audio recording. The first subtraction component determines a local linear transformation of the sound recording and subtracts the local linear transformation of the sound recording from the mixed audio recording to generate a new mixed audio recording. The second subtraction component compares one or more segments of the sound recording with one or more corresponding segments of the new mixed audio recording and reduces a power level of the new mixed audio recording based at least in part on correlation of the one or more corresponding segments with the one or more segments.

Abstract translation: 呈现用于从音频记录中去除声音记录的系统和技术（例如，嵌入在媒体文件中的音频记录）。 系统可以包括识别组件，第一减法组件和第二减法组件。 识别组件识别混合音频记录中的声音记录。 第一减法分量确定声音记录的局部线性变换，并从混合音频记录中减去声音记录的局部线性变换，以产生新的混合音频记录。 第二减法组件将声音录制的一个或多个段与新的混合音频记录的一个或多个对应的段进行比较，并且至少部分地基于一个或多个相应段的相关性来降低新的混合音频记录的功率水平 与一个或多个段。

公开(公告)号：US09971940B1

公开(公告)日：2018-05-15

申请号：US15231445

申请日：2016-08-08

Applicant: Google Inc.

Inventor： Luciano Sbaiz ,  Jay Yagnik ,  King Hong Thomas Leung ,  Hanna Pasula ,  Thomas Chadwick Walters ,  Thomas Bugnon ,  Matthias Rochus Konrad

IPC: G06K9/62 ,  G06K9/00 ,  G06F21/32 ,  G06N3/04

CPC classification number: G06K9/00758 ,  G06F21/10 ,  G06F2221/074 ,  G06K9/00718 ,  G06K9/00765 ,  G06K9/6215 ,  G06K9/6255 ,  G06K9/6269 ,  G06N3/04

Abstract: Provided content is determined to contain an asset represented by reference content by comparing digital fingerprints of the provided content and the reference content. The fingerprints of the reference content and the provided content are generated using a convolutional neural network (CNN). The CNN is trained using a plurality of frame triplets including an anchor frame representing the reference content, a positive frame which is a transformation of the anchor frame, and a negative frame representing content that is not the reference content. The provided content is determined to contain the asset represented by the reference content based on a similarity measure between the generated fingerprints. If the provided content is determined to contain the asset represented by the reference content, a policy associated with the asset is enforced on the provided content.

公开(公告)号：US20180018580A1

公开(公告)日：2018-01-18

申请号：US15212037

申请日：2016-07-15

Applicant: Google Inc.

Inventor： Benjamin Kenneth Coppin ,  Mustafa Suleyman ,  Thomas Chadwick Walters ,  Timothy Mann ,  Chia-Yueh Carlton Chu ,  Martin Szummer ,  Luis Carlos Cobo Rus ,  Jean-Francois Crespo

IPC: G06N99/00 ,  G06F17/30 ,  G06N7/00

CPC classification number: G06N20/00 ,  G06F16/24578 ,  G06F16/9535 ,  G06N3/0445 ,  G06N3/084 ,  G06N7/005

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using a machine learning model that has been trained through reinforcement learning to select a content item. One of the methods includes receiving first data characterizing a first context in which a first content item may be presented to a first user in a presentation environment; and providing the first data as input to a long-term engagement machine learning model, the model having been trained through reinforcement learning to: receive a plurality of inputs, and process each of the plurality of inputs to generate a respective engagement score for each input that represents a predicted, time-adjusted total number of selections by the respective user of future content items presented to the respective user in the presentation environment if the respective content item is presented in the respective context.

公开(公告)号：US09679579B1

公开(公告)日：2017-06-13

申请号：US15186993

申请日：2016-06-20

Applicant: Google Inc.

Inventor： Richard Francis Lyon ,  Ron Weiss ,  Thomas Chadwick Walters

IPC: G10L21/028 ,  G10L21/0356 ,  G10L25/51 ,  G10L21/0308 ,  G10L21/0388

CPC classification number: G10L21/0356 ,  G10L21/02 ,  G10L21/0272 ,  G10L21/028 ,  G10L21/0308 ,  G10L21/0388 ,  G10L25/51

Abstract: Systems and methods facilitating removal of content from audio files are described. A method includes identifying a sound recording in a first audio file, identifying a reference file having at least a defined level of similarity to the sound recording, and processing the first audio file to remove the sound recording and generate a second audio file. In some embodiments, winner-take-all coding and Hough transforms are employed for determining alignment and rate adjustment of the reference file in the first audio file. After alignment, the reference file is filtered in the frequency domain to increase similarity between the reference file and the sound recording. The frequency domain representation (FR) of the filtered version is subtracted from the FR first audio and the result converted to a time representation of the second audio file. In some embodiments, spectral subtraction is also performed to generate a further improved second audio file.

公开(公告)号：US09373320B1

公开(公告)日：2016-06-21

申请号：US13972262

申请日：2013-08-21

Applicant: Google Inc.

Inventor： Richard Francis Lyon ,  Ron Weiss ,  Thomas Chadwick Walters

IPC: G10L21/028 ,  G10L13/033 ,  G10L21/02 ,  G10L21/0272 ,  G10L21/0308

CPC classification number: G10L21/0356 ,  G10L21/02 ,  G10L21/0272 ,  G10L21/028 ,  G10L21/0308 ,  G10L21/0388 ,  G10L25/51

Abstract: Systems and methods facilitating removal of content from audio files are described. A method includes identifying a sound recording in a first audio file, identifying a reference file having at least a defined level of similarity to the sound recording, and processing the first audio file to remove the sound recording and generate a second audio file. In some embodiments, winner-take-all coding and Hough transforms are employed for determining alignment and rate adjustment of the reference file in the first audio file. After alignment, the reference file is filtered in the frequency domain to increase similarity between the reference file and the sound recording. The frequency domain representation (FR) of the filtered version is subtracted from the FR first audio and the result converted to a time representation of the second audio file. In some embodiments, spectral subtraction is also performed to generate a further improved second audio file.

Abstract translation: 描述了有助于从音频文件中移除内容的系统和方法。 一种方法包括：识别在第一音频文件中的声音记录，识别具有至少与所述声音记录的确定的相似度水平的参考文件，以及处理所述第一音频文件以去除所述声音记录并生成第二音频文件。 在一些实施例中，采用获胜者全部编码和霍夫变换来确定第一音频文件中的参考文件的对准和速率调整。 对齐后，参考文件在频域中进行过滤，以增加参考文件和录音之间的相似性。 从FR第一音频中减去滤波版本的频域表示（FR），并将结果转换为第二音频文件的时间表示。 在一些实施例中，还执行频谱减法以产生进一步改进的第二音频文件。

公开(公告)号：US09659014B1

公开(公告)日：2017-05-23

申请号：US14024150

申请日：2013-09-11

Applicant: Google Inc.

Inventor： Thomas Chadwick Walters ,  Gertjan Pieter Halkes ,  Matthias Rochus Konrad ,  Gheorghe Postelnicu

IPC: G06F17/30

CPC classification number: G06F17/30598 ,  G06F17/30026 ,  G06F17/30032 ,  G06F17/30454 ,  G06F17/3074 ,  G06F17/30743 ,  G06F17/30787 ,  G06F17/30799

Abstract: Aspects relate to determining whether a probe media content matches one or more reference media content. The reference media content is classified into a content class. The probe media content could also be classified into a content class. Similarities between the probe media content and the reference media content are identified. A matching score given to the probe media content is weighted based on statistics regarding matches and false-positive rates for the content class of the reference media content. Further, classifiers can be trained on computed audio features and video features and/or video metadata and audio metadata of various media content.

公开(公告)号：US20130338806A1

公开(公告)日：2013-12-19

申请号：US13730076

申请日：2012-12-28

Applicant: Google Inc.

Inventor： Christopher Russell LaRosa ,  Sam Kvaalen ,  Thomas Chadwick Walters ,  Richard Francis Lyon ,  Robert Steven Glickstein ,  Rushabh Ashok Doshi ,  Molly Castle Nix ,  Jason Matthew Toff

IPC: G06F3/16

CPC classification number: G06F3/16 ,  G10H1/0008 ,  G10H1/0025 ,  G10H2210/021 ,  G10H2210/125 ,  G10L21/0272 ,  G11B27/034 ,  G11B27/28 ,  H04H60/04 ,  Y02D70/1224 ,  Y02D70/1226 ,  Y02D70/1242 ,  Y02D70/1244 ,  Y02D70/1246 ,  Y02D70/1262 ,  Y02D70/1264 ,  Y02D70/142 ,  Y02D70/144 ,  Y02D70/146 ,  Y02D70/168

Abstract: Systems and techniques for removing a sound recording from an audio recording (e.g., an audio recording embedded in a media file) are presented. The system can include an identification component, a first subtraction component and a second subtraction component. The identification component identifies a sound recording in a mixed audio recording. The first subtraction component determines a local linear transformation of the sound recording and subtracts the local linear transformation of the sound recording from the mixed audio recording to generate a new mixed audio recording. The second subtraction component compares one or more segments of the sound recording with one or more corresponding segments of the new mixed audio recording and reduces a power level of the new mixed audio recording based at least in part on correlation of the one or more corresponding segments with the one or more segments.

Abstract translation: 呈现用于从音频记录中去除声音记录的系统和技术（例如，嵌入在媒体文件中的音频记录）。 系统可以包括识别组件，第一减法组件和第二减法组件。 识别组件识别混合音频记录中的声音记录。 第一减法分量确定声音记录的局部线性变换，并从混合音频记录中减去声音记录的局部线性变换，以产生新的混合音频记录。 第二减法组件将声音录制的一个或多个段与新的混合音频记录的一个或多个对应的段进行比较，并且至少部分地基于一个或多个相应段的相关性来降低新的混合音频记录的功率水平 与一个或多个段。

公开(公告)号：US20170256271A1

公开(公告)日：2017-09-07

申请号：US15600321

申请日：2017-05-19

Applicant: Google Inc.

Inventor： Richard Francis Lyon ,  Ron Weiss ,  Thomas Chadwick Walters

IPC: G10L21/0356 ,  G10L21/0308 ,  G10L25/51 ,  G10L21/0388

CPC classification number: G10L21/0356 ,  G10L21/02 ,  G10L21/0272 ,  G10L21/028 ,  G10L21/0308 ,  G10L21/0388 ,  G10L25/51 ,  G10L2021/02087

Abstract: Systems and methods facilitating removal of content from audio files are described. A method includes identifying a sound recording in a first audio file, identifying a reference file having at least a defined level of similarity to the sound recording, and processing the first audio file to remove the sound recording and generate a second audio file. In some embodiments, winner-take-all coding and Hough transforms are employed for determining alignment and rate adjustment of the reference file in the first audio file. After alignment, the reference file is filtered in the frequency domain to increase similarity between the reference file and the sound recording. The frequency domain representation (FR) of the filtered version is subtracted from the FR first audio and the result converted to a time representation of the second audio file. In some embodiments, spectral subtraction is also performed to generate a further improved second audio file.

公开(公告)号：US20160041807A1

公开(公告)日：2016-02-11

申请号：US14920463

申请日：2015-10-22

Applicant: Google Inc.

Inventor： Christopher Russell LaRosa ,  Sam Kvaalen ,  Thomas Chadwick Walters ,  Richard Francis Lyon ,  Robert Steven Glickstein ,  Rushabh Ashok Doshi ,  Molly Castle Nix ,  Jason Matthew Toff

IPC: G06F3/16 ,  G10L21/0272 ,  G10H1/00 ,  G11B27/034

Abstract: Systems and techniques for removing a sound recording from an audio recording (e.g., an audio recording embedded in a media file) are presented. The system can include an identification component, a first subtraction component and a second subtraction component. The identification component identifies a sound recording in a mixed audio recording. The first subtraction component determines a local linear transformation of the sound recording and subtracts the local linear transformation of the sound recording from the mixed audio recording to generate a new mixed audio recording. The second subtraction component compares one or more segments of the sound recording with one or more corresponding segments of the new mixed audio recording and reduces a power level of the new mixed audio recording based at least in part on correlation of the one or more corresponding segments with the one or more segments.