公开(公告)号：US20250157459A1

公开(公告)日：2025-05-15

申请号：US18838867

申请日：2022-03-24

Applicant: Yiming WANG ,  Chengyi WANG ,  Jinyu LI ,  Yu WU ,  Shujie LIU ,  MICROSOFT TECHNOLOGY LICENSING, LLC

Inventor： Yiming WANG ,  Chengyi WANG ,  Jinyu LI ,  Yu WU ,  Shujie LIU

IPC: G10L15/06 ,  G10L15/16

Abstract: Systems and methods are provided for generating a pseudo-labeled training dataset by at least one of: (1) extracting a set of intermediate outputs from an automatic speech recognition model based on applying the automatic speech recognition model to the set of unlabeled speech data, clustering the set of intermediate outputs into different clusters, and generating a first set of pseudo-labels comprising cluster assignments associated with the different clusters and which correspond to the unlabeled speech data, or (2) generating a set of decoded word sequences for the unlabeled speech data by applying the automatic speech recognition model to the set of unlabeled speech data, and generating a second set of pseudo-labels associated with the unlabeled speech data by applying the automatic speech recognition model to both (i) the set of decoded word sequences and (ii) the set of unlabeled speech data.

公开(公告)号：US08226844B2

公开(公告)日：2012-07-24

申请号：US13051893

申请日：2011-03-18

Applicant: Mingyuan Gao ,  Fengqin Hu ,  Shujie Liu ,  Xianyong Lu

Inventor： Mingyuan Gao ,  Fengqin Hu ,  Shujie Liu ,  Xianyong Lu

IPC: H01F1/33

CPC classification number: B82Y30/00 ,  A61K49/186 ,  B22F1/0018 ,  B22F9/30 ,  B22F2999/00 ,  B82Y5/00 ,  B82Y25/00 ,  H01F1/0054 ,  B22F1/0062

Abstract: This invention is concerned with biocompatible magnetic nanocrystals highly soluble and dispersible in a physiological buffer, powder of biocompatible magnetic nanocrystals and nanocrystals bearing surface reactive N-hydroxysuccinimide ester moiety, and preparations thereof. The magnetic nanocrystals in powder form are highly soluble in a physiological buffer. The resultant aqueous colloidal solution presents long term stability in ambient conditions. Moreover, the carboxyl group on the surface of the magnetic nanocrystals can be converted to N-hydroxysuccinimide ester moiety in an organic solvent. The resultant powder of the magnetic nanocrystals carrying surface N-hydroxysuccinimide ester moiety is soluble and dispersible in an aqueous solution. Different types of biomolecules bearing amino group can covalently be attached to the magnetic nanocrystal simply by mixing them in aqueous solutions. Moreover, the powder of the magnetic nanocrystals bearing surface N-hydroxysuccinimide ester moiety retain reaction activity with biomolecule after long-term storage.

Abstract translation: 本发明涉及在生理缓冲液，生物相容性磁性纳米晶体粉末和具有表面活性N-羟基琥珀酰亚胺酯部分的纳米晶体及其制剂中高度可溶和分散的生物相容性磁性纳米晶体。 粉末形式的磁性纳米晶体高度可溶于生理缓冲液。 所得水性胶体溶液在环境条件下具有长期稳定性。 此外，磁性纳米晶体表面上的羧基可以在有机溶剂中转化为N-羟基琥珀酰亚胺酯部分。 携带表面N-羟基琥珀酰亚胺酯部分的磁性纳米晶体的所得粉末是可溶的并可分散在水溶液中。 带有氨基的不同类型的生物分子可以简单地通过将它们混合在水溶液中来共价连接到磁性纳米晶体上。 此外，具有表面N-羟基琥珀酰亚胺酯部分的磁性纳米晶体的粉末在长期储存后保持与生物分子的反应活性。

公开(公告)号：US10198792B2

公开(公告)日：2019-02-05

申请号：US13502115

申请日：2010-10-14

Applicant: Wang Lin Lai ,  Shujie Liu ,  Dong Tian

Inventor： Wang Lin Lai ,  Shujie Liu ,  Dong Tian

IPC: H04N13/00 ,  G06T5/00 ,  H04N19/597 ,  H04N19/46 ,  H04N19/61 ,  H04N19/117 ,  H04N19/14 ,  H04N19/182 ,  G06T7/12 ,  H04N13/128

Abstract: Several implementations relate, for example, to depth encoding and/or filtering for 3D video (3DV) coding formats. A sparse dyadic mode for partitioning macroblocks (MBs) along edges in a depth map is provided as well as techniques for trilateral (or bilateral) filtering of depth maps that may include adaptive selection between filters sensitive to changes in video intensity and/or changes in depth. One implementation partitions a depth picture, and then refines the partitions based on a corresponding image picture. Another implementation filters a portion of a depth picture based on values for a range of pixels in the portion. For a given pixel in the portion that is being filtered, the filter weights a value of a particular pixel in the range by a weight that is based on one or more of location distance, depth difference, and image difference.

公开(公告)号：US09031357B2

公开(公告)日：2015-05-12

申请号：US13463934

申请日：2012-05-04

Applicant: Philip Andrew Chou ,  Cha Zhang ,  Zhengyou Zhang ,  Shujie Liu

Inventor： Philip Andrew Chou ,  Cha Zhang ,  Zhengyou Zhang ,  Shujie Liu

IPC: G06K9/32 ,  G06T7/00

CPC classification number: G06K9/32 ,  G06T7/593

Abstract: A temporal information integration dis-occlusion system and method for using historical data to reconstruct a virtual view containing an occluded area. Embodiments of the system and method use temporal information of the scene captured previously to obtain a total history. This total history is warped onto information captured by a camera at a current time in order to help reconstruct the dis-occluded areas. The historical data (or frames) from the total history match only a portion of the frames contained in the captured information. This warping yields warped history information. Warping is performed by using one of two embodiments to match points in an estimation of the current information to points in the captured information. Next, regions of current information are split using a classifier. The warped history information and the captured information then are merged to obtain an estimate for the current information and the reconstructed virtual view.

Abstract translation: 一种用于使用历史数据重建包含遮挡区域的虚拟视图的时间信息整合遮挡系统和方法。 系统和方法的实施例使用先前捕获的场景的时间信息来获得总历史。 这个总历史在当前时间由相机拍摄的信息扭曲，以帮助重建被遮挡的区域。 来自总历史记录的历史数据（或帧）仅匹配捕获信息中包含的帧的一部分。 这种扭曲产生扭曲的历史信息。 通过使用两个实施例中的一个实现扭曲，以将当前信息的估计中的点与捕获的信息中的点进行匹配。 接下来，使用分类器分割当前信息的区域。 然后将翘曲的历史信息和捕获的信息合并，以获得当前信息和重建的虚拟视图的估计。

公开(公告)号：US20240265924A1

公开(公告)日：2024-08-08

申请号：US18573846

申请日：2021-06-29

Applicant: Shujie LIU ,  Jinyu LI ,  Long ZHOU ,  Xie SUN ,  Microsoft Technology Licensing, LLC

Inventor： Jinyu LI ,  Long ZHOU ,  Xie SUN ,  Shujie LIU

IPC: G10L15/32 ,  G10L15/00 ,  G10L15/06 ,  G10L15/30

CPC classification number: G10L15/32 ,  G10L15/005 ,  G10L15/063 ,  G10L15/30 ,  G10L2015/0635

Abstract: Embodiments are provided for building a configurable multilingual model. A computing system obtains a plurality of language-specific automatic speech recognition modules and a universal automatic speech recognition module trained on a multi-language training dataset comprising training data corresponding to each of the plurality of different languages. The computing system then compiles the universal automatic speech recognition module with the plurality of language-specific automatic speech recognition modules to generate a configurable multilingual model that is configured to selectively and dynamically utilize a sub-set of the plurality of language-specific automatic speech recognition modules with the universal automatic speech recognition module to process audio content in response to user input identifying one or more target languages associated with the audio content.

公开(公告)号：US20140184744A1

公开(公告)日：2014-07-03

申请号：US14239918

申请日：2011-08-26

Applicant: Shujie Liu ,  Wang Lin Lai ,  Dong Tian

Inventor： Shujie Liu ,  Wang Lin Lai ,  Dong Tian

IPC: H04N19/597 ,  H04N19/167

Abstract: Various implementations address depth coding and related disciplines. In one particular implementation, a segmentation is determined for a particular portion of a video image in a sequence of video images. The segmentation is determined based on reference depth indicators that are associated with at least a portion of one video image in the sequence of video images. Target depth indicators associated with the particular portion of the video image are processed. The processing is based on the determined segmentation in the particular portion of the video image. In another particular implementation, a segmentation is determined for at least a given portion of a video image based on depth indicators associated with the given portion. The segmentation is extended from the given portion into a target portion of the video image based on pixel values in the given portion and on pixel values in the target portion.

Abstract translation: 各种实现涉及深度编码和相关学科。 在一个特定实现中，为视频图像序列中的视频图像的特定部分确定分割。 基于与视频图像序列中的一个视频图像的至少一部分相关联的参考深度指示符来确定分割。 处理与视频图像的特定部分相关联的目标深度指示符。 该处理基于视频图像的特定部分中确定的分割。 在另一特定实现中，基于与给定部分相关联的深度指示符，确定视频图像的至少一个给定部分的分割。 基于给定部分中的像素值和目标部分中的像素值，将分割从给定部分扩展到视频图像的目标部分。

公开(公告)号：US08017031B2

公开(公告)日：2011-09-13

申请号：US12008864

申请日：2008-01-14

Applicant: Mingyuan Gao ,  Fengqin Hu ,  Shujie Liu ,  Xianyong Lu

Inventor： Mingyuan Gao ,  Fengqin Hu ,  Shujie Liu ,  Xianyong Lu

IPC: H01F1/33

CPC classification number: B82Y30/00 ,  A61K49/186 ,  B22F1/0018 ,  B22F9/30 ,  B22F2999/00 ,  B82Y5/00 ,  B82Y25/00 ,  H01F1/0054 ,  B22F1/0062

Abstract: This invention is concerned with biocompatible magnetic nanocrystals highly soluble and dispersible in a physiological buffer, powder of biocompatible magnetic nanocrystals and nanocrystals bearing surface reactive N-hydroxysuccinimide ester moiety, and preparations thereof. The magnetic nanocrystals in powder form are highly soluble in a physiological buffer. The resultant aqueous colloidal solution presents long term stability in ambient conditions. Moreover, the carboxyl group on the surface of the magnetic nanocrystals can be converted to N-hydroxysuccinimide ester moiety in an organic solvent. The resultant powder of the magnetic nanocrystals carrying surface N-hydroxysuccinimide ester moiety is soluble and dispersible in an aqueous solution. Different types of biomolecules bearing amino group can covalently be attached to the magnetic nanocrystal simply by mixing them in aqueous solutions. Moreover, the powder of the magnetic nanocrystals bearing surface N-hydroxysuccinimide ester moiety retain reaction activity with biomolecule after long-term storage.

Abstract translation: 本发明涉及在生理缓冲液，生物相容性磁性纳米晶体粉末和具有表面活性N-羟基琥珀酰亚胺酯部分的纳米晶体及其制剂中高度可溶和分散的生物相容性磁性纳米晶体。 粉末形式的磁性纳米晶体高度可溶于生理缓冲液。 所得水性胶体溶液在环境条件下具有长期稳定性。 此外，磁性纳米晶体表面上的羧基可以在有机溶剂中转化为N-羟基琥珀酰亚胺酯部分。 携带表面N-羟基琥珀酰亚胺酯部分的磁性纳米晶体的所得粉末是可溶的并可分散在水溶液中。 带有氨基的不同类型的生物分子可以简单地通过将它们混合在水溶液中来共价连接到磁性纳米晶体上。 此外，具有表面N-羟基琥珀酰亚胺酯部分的磁性纳米晶体的粉末在长期储存后保持与生物分子的反应活性。

公开(公告)号：US12249336B2

公开(公告)日：2025-03-11

申请号：US18573846

申请日：2021-06-29

Applicant: Microsoft Technology Licensing, LLC ,  Jinyu Li ,  Long Zhou ,  Xie Sun ,  Shujie Liu

Inventor： Jinyu Li ,  Long Zhou ,  Xie Sun ,  Shujie Liu

IPC: G10L15/32 ,  G10L15/00 ,  G10L15/06 ,  G10L15/30

Abstract: Embodiments are provided for building a configurable multilingual model. A computing system obtains a plurality of language-specific automatic speech recognition modules and a universal automatic speech recognition module trained on a multi-language training dataset comprising training data corresponding to each of the plurality of different languages. The computing system then compiles the universal automatic speech recognition module with the plurality of language-specific automatic speech recognition modules to generate a configurable multilingual model that is configured to selectively and dynamically utilize a sub-set of the plurality of language-specific automatic speech recognition modules with the universal automatic speech recognition module to process audio content in response to user input identifying one or more target languages associated with the audio content.

公开(公告)号：US20130294710A1

公开(公告)日：2013-11-07

申请号：US13463934

申请日：2012-05-04

Applicant: Philip Andrew Chou ,  Cha Zhang ,  Zhengyou Zhang ,  Shujie Liu

Inventor： Philip Andrew Chou ,  Cha Zhang ,  Zhengyou Zhang ,  Shujie Liu

IPC: G06K9/32

CPC classification number: G06K9/32 ,  G06T7/593

Abstract: A temporal information integration dis-occlusion system and method for using historical data to reconstruct a virtual view containing an occluded area. Embodiments of the system and method use temporal information of the scene captured previously to obtain a total history. This total history is warped onto information captured by a camera at a current time in order to help reconstruct the dis-occluded areas. The historical data (or frames) from the total history match only a portion of the frames contained in the captured information. This warping yields warped history information. Warping is performed by using one of two embodiments to match points in an estimation of the current information to points in the captured information. Next, regions of current information are split using a classifier. The warped history information and the captured information then are merged to obtain an estimate for the current information and the reconstructed virtual view.

Abstract translation: 一种用于使用历史数据重建包含遮挡区域的虚拟视图的时间信息整合遮挡系统和方法。 系统和方法的实施例使用先前捕获的场景的时间信息来获得总历史。 这个总历史在当前时间由相机拍摄的信息扭曲，以帮助重建被遮挡的区域。 来自总历史记录的历史数据（或帧）仅匹配捕获信息中包含的帧的一部分。 这种扭曲产生扭曲的历史信息。 通过使用两个实施例中的一个实现扭曲，以将当前信息的估计中的点与捕获的信息中的点进行匹配。 接下来，使用分类器分割当前信息的区域。 然后将翘曲的历史信息和捕获的信息合并，以获得当前信息和重建的虚拟视图的估计。

公开(公告)号：US20120200669A1

公开(公告)日：2012-08-09

申请号：US13502115

申请日：2010-10-14

Applicant: Wang Lin Lai ,  Shujie Liu ,  Dong Tian

Inventor： Wang Lin Lai ,  Shujie Liu ,  Dong Tian

IPC: H04N13/00

CPC classification number: G06T5/002 ,  G06T7/12 ,  G06T2207/10028 ,  G06T2207/20016 ,  G06T2207/20028 ,  H04N13/128 ,  H04N19/117 ,  H04N19/14 ,  H04N19/182 ,  H04N19/46 ,  H04N19/597 ,  H04N19/61 ,  H04N2213/003

Abstract: Several implementations relate, for example, to depth encoding and/or filtering for 3D video (3DV) coding formats. A sparse dyadic mode for partitioning macroblocks (MBs) along edges in a depth map is provided as well as techniques for trilateral (or bilateral) filtering of depth maps that may include adaptive selection between filters sensitive to changes in video intensity and/or changes in depth. One implementation partitions a depth picture, and then refines the partitions based on a corresponding image picture. Another implementation filters a portion of a depth picture based on values for a range of pixels in the portion. For a given pixel in the portion that is being filtered, the filter weights a value of a particular pixel in the range by a weight that is based on one or more of location distance, depth difference, and image difference.

Abstract translation: 几个实现涉及例如3D视频（3DV）编码格式的深度编码和/或滤波。 提供了用于在深度图中沿着边缘分割宏块（MB）的稀疏二元模式以及用于深度图的三边（或双边）滤波的技术，其可以包括对视频强度的变化敏感的滤波器和/或 深度。 一个实现分割深度图片，然后基于相应的图像图像来精化分区。 另一个实施方式基于该部分中的像素范围的值来过滤一部分深度图像。 对于被滤波的部分中的给定像素，滤波器将基于位置距离，深度差和图像差异中的一个或多个的权重来对范围中的特定像素的值进行加权。