公开(公告)号：US10277919B2

公开(公告)日：2019-04-30

申请号：US15077232

申请日：2016-03-22

Applicant: GOOGLE INC.

Inventor： Anil Kokaram ,  Damien Kelly ,  Andrew Joseph Crawford ,  Hugh Pierre Denman

IPC: H04N19/65 ,  H04N19/80 ,  H04N19/85 ,  G06T5/40 ,  H04N5/14 ,  H04N19/117 ,  H04N19/142 ,  H04N19/179 ,  H04N19/86 ,  G06T5/00

Abstract: Described herein are techniques related to noise reduction for image sequences or videos. This Abstract is submitted with the understanding that it will not be used to interpret or limit the scope and meaning of the claims. A noise reduction tool includes a motion estimator configured to estimated motion in the video, a noise spectrum estimator configured to estimate noise in the video, a shot detector configured to trigger the noise estimation process, a noise spectrum validator configured to validate the estimated noise spectrum, and a noise reducer to reduce noise in the video using the estimated noise spectrum.

公开(公告)号：US09524733B2

公开(公告)日：2016-12-20

申请号：US13891978

申请日：2013-05-10

Applicant: Google Inc.

Inventor： Jan Skoglund ,  Andrew J. Hines ,  Noami A. Harte ,  Anil Kokaram

IPC: G10L25/60

CPC classification number: G10L25/60

Abstract: Methods and systems are provided for using a model of human speech quality perception to provide an objective measure for predicting subjective quality assessments. A Virtual Speech Quality Objective Listener (ViSQOL) model is a signal-based full-reference metric that uses a spectro-temporal measure of similarity between a reference signal and test speech signal. Specifically, the model provides for the ability to detect and predict the level of clock drift, and determine whether such clock drift will impact a listener's quality of experience.

Abstract translation: 提供了使用人类语言质量感知模型的方法和系统来提供用于预测主观质量评估的客观量度。 虚拟语音质量目标监听器（ViSQOL）是一种基于信号的全参考度量，它使用参考信号和测试语音信号之间的相似性的频谱测量。 具体来说，该模型提供了检测和预测时钟漂移水平的能力，并确定这种时钟漂移是否会影响听众的体验质量。

公开(公告)号：US09263061B2

公开(公告)日：2016-02-16

申请号：US13899381

申请日：2013-05-21

Applicant: Google Inc.

Inventor： Andrew J. Hines ,  Jan Skoglund ,  Naomi Harte ,  Anil Kokaram

IPC: G10L25/78

CPC classification number: G10L25/78 ,  G10L21/0232 ,  G10L25/60

Abstract: Methods and systems are provided for detecting chop in an audio signal. A time-frequency representation, such as a spectrogram, is created for an audio signal and used to calculate a gradient of mean power per frame of the audio signal. Positive and negative gradients are defined for the signal based on the gradient of mean power, and a maximum overlap offset between the positive and negative gradients is determined by calculating a value that maximizes the cross-correlation of the positive and negative gradients. The negative gradient values may be combined (e.g., summed) with the overlap offset, and the combined values then compared with a threshold to estimate the amount of chop present in the audio signal. The chop detection model provided is low-complexity and is applicable to narrowband, wideband, and superwideband speech.

Abstract translation: 提供了用于检测音频信号中的斩波的方法和系统。 为音频信号创建时频表示，如频谱图，用于计算音频信号每帧平均功率的梯度。 基于平均功率梯度的信号定义正和负梯度，通过计算使正和负梯度的互相关最大化的值来确定正梯度和负梯度之间的最大重叠偏移。 负梯度值可以与重叠偏移组合（例如，相加），然后将组合值与阈值进行比较以估计音频信号中存在的斩波量。 提供的斩波检测模型是低复杂度的，适用于窄带，宽带和超宽带语音。

公开(公告)号：US10674045B2

公开(公告)日：2020-06-02

申请号：US15609826

申请日：2017-05-31

Applicant: Google Inc.

Inventor： Neil Birkbeck ,  Mohammad Izadi ,  Anil Kokaram ,  Balineedu C. Adsumilli ,  Damien Kelly

IPC: G06K9/00 ,  H04N5/21 ,  G06T5/00

Abstract: Implementations disclose mutual noise estimation for videos. A method includes determining an optimal frame noise variance for intensity values of each frame of frames of a video, the optimal frame noise variance based on a determined relationship between spatial variance and temporal variance of the intensity values of homogeneous blocks in the frame, identifying an optimal video noise variance for the video based on optimal frame noise variances of the frames of the video, selecting, for each frame of the video, one or more of the blocks having a spatial variance that is less than the optimal video noise variance, the one or more frames selected as the homogeneous blocks, and utilizing the selected homogeneous blocks to estimate a noise signal of the video.

公开(公告)号：US20140328548A1

公开(公告)日：2014-11-06

申请号：US13907890

申请日：2013-06-01

Applicant: Google Inc.

Inventor： Thor Carpenter ,  Anil Kokaram

IPC: G06T5/00

CPC classification number: G06T5/007 ,  G06T7/194 ,  G06T2207/10016 ,  G06T2207/30201

Abstract: Implementations generally relate to enhancing content appearance. In some implementations, a method includes receiving an image, selecting a reference object in the image. The method also includes determining one or more image parameter adjustments based on the selected reference object, and applying the one or more image parameter adjustments to the entire image.

Abstract translation: 实现通常涉及增强内容外观。 在一些实现中，一种方法包括接收图像，在图像中选择参考对象。 该方法还包括基于所选择的参考对象来确定一个或多个图像参数调整，以及将一个或多个图像参数调整应用于整个图像。

公开(公告)号：US10096114B1

公开(公告)日：2018-10-09

申请号：US14092413

申请日：2013-11-27

Applicant: Google Inc.

Inventor： Joan Lasenby ,  Stuart Bennett ,  Sasi Inguva ,  Damien Kelly ,  Andrew Crawford ,  Hugh Denman ,  Anil Kokaram

IPC: G06T7/00 ,  H04N17/00

Abstract: A method for determining the position of multiple cameras relative to each other includes at a processor, receiving video data from at least one video recording taken by each camera; selecting a subset of frames of each video recording, including determining relative blurriness of each frame of each video recording, selecting frames having a lowest relative blurriness, counting features points in each of the lowest relative blurriness frames, and selecting for further analysis, lowest relative blurriness frames having a highest count of feature points; and processing each selected subset of frames from each video recording to estimate the location and orientation of each camera.

公开(公告)号：US08908102B1

公开(公告)日：2014-12-09

申请号：US14017122

申请日：2013-09-03

Applicant: Google Inc.

Inventor： Andrew Joseph Crawford ,  Damien Kelly ,  Anil Kokaram ,  Hugh Pierre Denman

IPC: H04N5/14 ,  H04N5/262

CPC classification number: H04N5/23267 ,  G06T7/30 ,  G06T2207/10016 ,  H04N5/23254 ,  H04N5/23296 ,  H04N5/2628

Abstract: A system for video stabilization is provided. The system includes a media component, a transformation component, an offset component and a zoom component. The media component receives a video sequence including at least a first video frame and a second video frame. The transformation component calculates at least a first motion parameter associated with translational motion for the first video frame and at least a second motion parameter associated with the translational motion for the second video frame. The offset component subtracts an offset value generated as a function of a maximum motion parameter and a minimum motion parameter from the first motion parameter and the second motion parameter to generate a set of modified motion parameters. The zoom component determines a zoom value for the video sequence based at least in part on the set of modified motion parameters.

Abstract translation: 提供了一种用于视频稳定的系统。 系统包括媒体组件，变换组件，偏移组件和缩放组件。 媒体组件接收包括至少第一视频帧和第二视频帧的视频序列。 变换分量至少计算与第一视频帧的平移运动相关联的第一运动参数和与第二视频帧的平移运动相关联的至少第二运动参数。 偏移分量从第一运动参数和第二运动参数减去作为最大运动参数和最小运动参数的函数产生的偏移值，以生成一组修改的运动参数。 缩放组件至少部分地基于修改的运动参数的集合来确定视频序列的缩放值。

公开(公告)号：US10454987B2

公开(公告)日：2019-10-22

申请号：US15337806

申请日：2016-10-28

Applicant: Google Inc.

Inventor： Chao Chen ,  Yao-Chung Lin ,  Anil Kokaram ,  Steve Benting

IPC: H04L29/06 ,  H04L12/811 ,  H04N21/2343 ,  H04N21/258

Abstract: Implementations disclose bitrate optimization for multi-representation encoding using playback statistics. A method includes generating multiple versions of a segment of a source video, the versions comprising encodings of the segment at different encoding bitrates for each resolution of the segment, measuring a quality metric for each version of the segment, generating rate-quality models for each resolution of the segment based on the measured quality metrics corresponding to the resolutions, generating a probability model to predict requesting probabilities that representations of the segment are requested, the probability model based on a joint probability distribution of network speed and viewport size that is generated from client-side feedback statistics associated with prior playbacks of other videos, determining an encoding bitrate for each of the representations of the segment based on the rate-quality models and the probability model, and assigning determined encoding bitrates to corresponding representations of the segment.

公开(公告)号：US20180352118A1

公开(公告)日：2018-12-06

申请号：US15609826

申请日：2017-05-31

Applicant: Google Inc.

Inventor： Neil Birkbeck ,  Mohammad Izadi ,  Anil Kokaram ,  Balineedu C. Adsumilli ,  Damien Kelly

IPC: H04N5/21

Abstract: Implementations disclose mutual noise estimation for videos. A method includes determining an optimal frame noise variance for intensity values of each frame of frames of a video, the optimal frame noise variance based on a determined relationship between spatial variance and temporal variance of the intensity values of homogeneous blocks in the frame, identifying an optimal video noise variance for the video based on optimal frame noise variances of the frames of the video, selecting, for each frame of the video, one or more of the blocks having a spatial variance that is less than the optimal video noise variance, the one or more frames selected as the homogeneous blocks, and utilizing the selected homogeneous blocks to estimate a noise signal of the video.

公开(公告)号：US09888255B1

公开(公告)日：2018-02-06

申请号：US15079251

申请日：2016-03-24

Applicant: GOOGLE INC.

Inventor： Anil Kokaram ,  Damien Kelly ,  Andrew Joseph Crawford

IPC: H04N19/577 ,  H04N19/59 ,  H04N19/587 ,  H04N19/52 ,  H04N19/527 ,  H04N19/597 ,  H04N19/53 ,  H04N19/194 ,  H04N19/553 ,  H04N19/44

CPC classification number: H04N19/577 ,  H04N7/0127 ,  H04N7/014 ,  H04N19/194 ,  H04N19/44 ,  H04N19/52 ,  H04N19/527 ,  H04N19/53 ,  H04N19/553 ,  H04N19/587 ,  H04N19/59 ,  H04N19/597

Abstract: A method for pull frame interpolation includes receiving an encoded bitstream including information representing a plurality of frames of video data, decoding the plurality of frames, including identifying a plurality of motion vectors indicating motion from a first frame of the plurality of video frames to a second frame of the plurality of video frames, identifying an interpolation point between the first frame and the second frame, identifying a plurality of candidate interpolation motion vectors indicating motion from the first frame to the interpolation point and from the second frame to the interpolation point based on the plurality of motion vectors, selecting an interpolation motion vector from the plurality of candidate interpolation motion vectors based on a metric, and generating an interpolated frame at the interpolation point based on the selected interpolation motion vector, which may include correcting an artifact in the interpolated frame by blending the interpolated frame.