公开(公告)号：US20080066136A1

公开(公告)日：2008-03-13

申请号：US11509250

申请日：2006-08-24

申请人： Chitra Dorai ,  Robert G. Farrell ,  Ying Li ,  Youngja Park

发明人： Chitra Dorai ,  Robert G. Farrell ,  Ying Li ,  Youngja Park

IPC分类号： G06F3/00 ,  G06F13/00 ,  H04N7/16 ,  H04N5/445

CPC分类号： H04N5/147 ,  G06F16/7834 ,  G06F16/7844

摘要： Computer implemented method, system and computer usable program code for detecting topic shift boundaries in a multimedia stream. A computer implemented method for detecting topic shift boundaries in a multimedia stream includes receiving a multimedia stream, and performing multimodal analysis on the multimedia stream to locate a plurality of temporal positions within the multimedia stream at which topic changes have an increased likelihood of occurring to provide a sequence of multimedia portions. Characteristics for a sliding window for each multimedia portion in the sequence of multimedia portions are automatically determined, and topic shift boundaries are detected in each multimedia portion by applying a text-based topic shift detector over the media stream's text transcript using a sliding window, wherein the sliding window used with each multimedia portion has the characteristics determined from its respective multimedia portion.

摘要翻译： 用于检测多媒体流中的主题移位边界的计算机实现的方法，系统和计算机可用程序代码。 一种用于检测多媒体流中的主题移位边界的计算机实现方法，包括：接收多媒体流，以及对所述多媒体流执行多模态分析，以定位所述多媒体流内的多个时间位置，在该多媒体流中，主题变化具有增加的发生可能性以提供 多媒体部分的序列。 自动确定多媒体部分序列中的每个多媒体部分的滑动窗口的特征，并且通过使用滑动窗口在媒体流的文本转录本上应用基于文本的主题移位检测器，在每个多媒体部分中检测主题移位边界，其中 与每个多媒体部分一起使用的滑动窗口具有从其各自的多媒体部分确定的特征。

公开(公告)号：US07382933B2

公开(公告)日：2008-06-03

申请号：US11210305

申请日：2005-08-24

申请人： Chitra Dorai ,  Ying Li ,  Youngja Park

发明人： Chitra Dorai ,  Ying Li ,  Youngja Park

IPC分类号： G06K9/36

CPC分类号： G06F17/30787 ,  G06F17/30796

摘要： System and method for partitioning a video into a series of semantic units where each semantic unit relates to a generally complete thematic topic. A computer implemented method for partitioning a video into a series of semantic units wherein each semantic unit relates to a theme or a topic, comprises dividing a video into a plurality of homogeneous segments, analyzing audio and visual content of the video, extracting a plurality of keywords from the speech content of each of the plurality of homogeneous segments of the video, and detecting and merging a plurality of groups of semantically related and temporally adjacent homogeneous segments into a series of semantic units in accordance with the results of both the audio and visual analysis and the keyword extraction. The present invention can be applied to generate important table-of-contents as well as index tables for videos to facilitate efficient video topic searching and browsing.

摘要翻译： 将视频分割成一系列语义单元的系统和方法，其中每个语义单元涉及一般完整的主题。 一种用于将视频分割成一系列语义单元的计算机实现的方法，其中每个语义单元涉及主题或主题，包括将视频划分为多个同构段，分析视频的音频和视觉内容，提取多个 根据视频的多个同构段的每个的语音内容的关键字，以及根据音频和视频的结果检测和合并多个语义相关和时间上相邻的同构段的组成一系列语义单元 分析和关键词提取。 本发明可以应用于产生重要的内容表以及用于视频的索引表，以便于有效的视频主题搜索和浏览。

公开(公告)号：US20070055695A1

公开(公告)日：2007-03-08

申请号：US11210305

申请日：2005-08-24

申请人： Chitra Dorai ,  Ying Li ,  Youngja Park

发明人： Chitra Dorai ,  Ying Li ,  Youngja Park

IPC分类号： G06F17/00

CPC分类号： G06F17/30787 ,  G06F17/30796

摘要： System and method for partitioning a video into a series of semantic units where each semantic unit relates to a generally complete thematic topic. A computer implemented method for partitioning a video into a series of semantic units wherein each semantic unit relates to a theme or a topic, comprises dividing a video into a plurality of homogeneous segments, analyzing audio and visual content of the video, extracting a plurality of keywords from the speech content of each of the plurality of homogeneous segments of the video, and detecting and merging a plurality of groups of semantically related and temporally adjacent homogeneous segments into a series of semantic units in accordance with the results of both the audio and visual analysis and the keyword extraction. The present invention can be applied to generate important table-of-contents as well as index tables for videos to facilitate efficient video topic searching and browsing.

摘要翻译： 将视频分割成一系列语义单元的系统和方法，其中每个语义单元涉及一般完整的主题。 一种用于将视频分割成一系列语义单元的计算机实现的方法，其中每个语义单元涉及主题或主题，包括将视频划分为多个同构段，分析视频的音频和视觉内容，提取多个 根据视频的多个同构段的每个的语音内容的关键字，以及根据音频和视频的结果检测和合并多个语义相关和时间上相邻的同构段的组成一系列语义单元 分析和关键词提取。 本发明可以应用于产生重要的内容表以及用于视频的索引表，以便于有效的视频主题搜索和浏览。

公开(公告)号：US08121432B2

公开(公告)日：2012-02-21

申请号：US12055023

申请日：2008-03-25

申请人： Chitra Dorai ,  Ying Li ,  Youngja Park

发明人： Chitra Dorai ,  Ying Li ,  Youngja Park

IPC分类号： G06K9/36

CPC分类号： G06F17/30787 ,  G06F17/30796

摘要： System and method for partitioning a video into a series of semantic units where each semantic unit relates to a generally complete thematic topic. A computer implemented method for partitioning a video into a series of semantic units wherein each semantic unit relates to a theme or a topic, comprises dividing a video into a plurality of homogeneous segments, analyzing audio and visual content of the video, extracting a plurality of keywords from the speech content of each of the plurality of homogeneous segments of the video, and detecting and merging a plurality of groups of semantically related and temporally adjacent homogeneous segments into a series of semantic units in accordance with the results of both the audio and visual analysis and the keyword extraction. The present invention can be applied to generate important table-of-contents as well as index tables for videos to facilitate efficient video topic searching and browsing.

摘要翻译： 将视频分割成一系列语义单元的系统和方法，其中每个语义单元涉及一般完整的主题。 一种用于将视频分割成一系列语义单元的计算机实现的方法，其中每个语义单元涉及主题或主题，包括将视频划分为多个同构段，分析视频的音频和视觉内容，提取多个 根据视频的多个同构段的每个的语音内容的关键字，以及根据音频和视频的结果检测和合并多个语义相关和时间上相邻的同构段的组成一系列语义单元 分析和关键词提取。 本发明可以应用于产生重要的内容表以及用于视频的索引表，以便于有效的视频主题搜索和浏览。

公开(公告)号：US20080175556A1

公开(公告)日：2008-07-24

申请号：US12055023

申请日：2008-03-25

申请人： Chitra Dorai ,  Ying Li ,  Youngja Park

发明人： Chitra Dorai ,  Ying Li ,  Youngja Park

IPC分类号： H04N5/93

CPC分类号： G06F17/30787 ,  G06F17/30796

摘要： System and method for partitioning a video into a series of semantic units where each semantic unit relates to a generally complete thematic topic. A computer implemented method for partitioning a video into a series of semantic units wherein each semantic unit relates to a theme or a topic, comprises dividing a video into a plurality of homogeneous segments, analyzing audio and visual content of the video, extracting a plurality of keywords from the speech content of each of the plurality of homogeneous segments of the video, and detecting and merging a plurality of groups of semantically related and temporally adjacent homogeneous segments into a series of semantic units in accordance with the results of both the audio and visual analysis and the keyword extraction. The present invention can be applied to generate important table-of-contents as well as index tables for videos to facilitate efficient video topic searching and browsing.

摘要翻译： 将视频分割成一系列语义单元的系统和方法，其中每个语义单元涉及一般完整的主题。 一种用于将视频分割成一系列语义单元的计算机实现的方法，其中每个语义单元涉及主题或主题，包括将视频划分为多个同构段，分析视频的音频和视觉内容，提取多个 根据视频的多个同构段的每个的语音内容的关键字，以及根据音频和视频的结果检测和合并多个语义相关和时间上相邻的同构段的组成一系列语义单元 分析和关键词提取。 本发明可以应用于产生重要的内容表以及用于视频的索引表，以便于有效的视频主题搜索和浏览。

公开(公告)号：US07668371B2

公开(公告)日：2010-02-23

申请号：US12126262

申请日：2008-05-23

申请人： Chitra Dorai ,  Ying Li

发明人： Chitra Dorai ,  Ying Li

IPC分类号： G06K9/34

CPC分类号： G06K9/00624

摘要： System and method for distinguishing between foreground content and background content in an image presentation. An initial background model is provided, and a final background model is constructed from the initial background model using the image presentation. The foreground content and background content in the image presentation are then distinguished from one another using the final background model. The present invention permits foreground content and background content to be separated from one another for further processing in different types of computer-generated image presentations such as digital slide presentations, video presentations, Web page presentations, and the like.

摘要翻译： 用于区分图像呈现中的前景内容和背景内容的系统和方法。 提供初始背景模型，并使用图像呈现从最初的背景模型构建最终背景模型。 然后使用最终背景模型将图像呈现中的前景内容和背景内容彼此区分开。 本发明允许前景内容和背景内容彼此分离，以便在不同类型的计算机生成的图像呈现中进行进一步处理，例如数字幻灯片呈现，视频呈现，网页呈现等。

公开(公告)号：US20080219554A1

公开(公告)日：2008-09-11

申请号：US12126262

申请日：2008-05-23

申请人： Chitra Dorai ,  Ying Li

发明人： Chitra Dorai ,  Ying Li

IPC分类号： G06K9/34

CPC分类号： G06K9/00624

摘要： System and method for distinguishing between foreground content and background content in an image presentation. An initial background model is provided, and a final background model is constructed from the initial background model using the image presentation. The foreground content and background content in the image presentation are then distinguished from one another using the final background model. The present invention permits foreground content and background content to be separated from one another for further processing in different types of computer-generated image presentations such as digital slide presentations, video presentations, Web page presentations, and the like.

摘要翻译： 用于区分图像呈现中的前景内容和背景内容的系统和方法。 提供初始背景模型，并使用图像呈现从最初的背景模型构建最终背景模型。 然后使用最终背景模型将图像呈现中的前景内容和背景内容彼此区分开。 本发明允许前景内容和背景内容彼此分离，以便在不同类型的计算机生成的图像呈现中进行进一步处理，例如数字幻灯片呈现，视频呈现，网页呈现等。

公开(公告)号：US20060153448A1

公开(公告)日：2006-07-13

申请号：US11034583

申请日：2005-01-13

申请人： Chitra Dorai ,  Ying Li

发明人： Chitra Dorai ,  Ying Li

IPC分类号： G06K9/34

CPC分类号： G06K9/00624

摘要： System and method for distinguishing between foreground content and background content in an image presentation. An initial background model is provided, and a final background model is constructed from the initial background model using the image presentation. The foreground content and background content in the image presentation are then distinguished from one another using the final background model. The present invention permits foreground content and background content to be separated from one another for further processing in different types of computer-generated image presentations such as digital slide presentations, video presentations, Web page presentations, and the like.

摘要翻译： 用于区分图像呈现中的前景内容和背景内容的系统和方法。 提供初始背景模型，并使用图像呈现从最初的背景模型构建最终背景模型。 然后使用最终背景模型将图像呈现中的前景内容和背景内容彼此区分开。 本发明允许前景内容和背景内容彼此分离，以便在不同类型的计算机生成的图像呈现中进行进一步处理，例如数字幻灯片演示，视频呈现，网页呈现等。

公开(公告)号：US20050286865A1

公开(公告)日：2005-12-29

申请号：US10892637

申请日：2004-07-16

申请人： Chitra Dorai ,  Ying Li

发明人： Chitra Dorai ,  Ying Li

IPC分类号： H04N5/783 ,  H04N5/85 ,  H04N7/24 ,  H04N9/804

CPC分类号： H04N21/4325 ,  H04N5/783 ,  H04N5/85 ,  H04N9/8042 ,  H04N21/440281 ,  H04N21/472 ,  H04N21/8455

摘要： Disclosed is a general framework for extracting semantics from composite media content at various resolutions. Specifically, given a media stream, which may consist of various types of media modalities including audio, visual, text and graphics information, the disclosed framework describes how various types of semantics could be extracted at different levels by exploiting and integrating different media features. The output of this framework is a series of tagged (or annotated) media segments at different scales. Specifically, at the lowest resolution, the media segments are characterized in a more general and broader sense, thus they are identified at a larger scale; while at the highest resolution, the media content is more specifically analyzed, inspected and identified, which thus results in small-scaled media segments.

摘要翻译： 公开了从复合媒体内容以各种分辨率提取语义的一般框架。 具体地，给定媒体流，其可以由包括音频，视觉，文本和图形信息的各种类型的媒体模式组成，所公开的框架描述了如何通过利用和集成不同媒体特征来在不同级别提取各种类型的语义。 该框架的输出是不同尺度的一系列带标签（或注释）的媒体片段。 具体来说，在最低分辨率下，媒体部分的特征更广泛和更广泛，因此它们被更广泛地识别出来。 而在最高分辨率下，更具体地分析，检查和识别媒体内容，从而导致小规模媒体片段。

公开(公告)号：US07890327B2

公开(公告)日：2011-02-15

申请号：US10892637

申请日：2004-07-16

申请人： Chitra Dorai ,  Ying Li

发明人： Chitra Dorai ,  Ying Li

IPC分类号： G10L15/14 ,  G10L15/06 ,  G10L15/00

CPC分类号： H04N21/4325 ,  H04N5/783 ,  H04N5/85 ,  H04N9/8042 ,  H04N21/440281 ,  H04N21/472 ,  H04N21/8455

摘要： Disclosed is a general framework for extracting semantics from composite media content at various resolutions. Specifically, given a media stream, which may consist of various types of media modalities including audio, visual, text and graphics information, the disclosed framework describes how various types of semantics could be extracted at different levels by exploiting and integrating different media features. The output of this framework is a series of tagged (or annotated) media segments at different scales. Specifically, at the lowest resolution, the media segments are characterized in a more general and broader sense, thus they are identified at a larger scale; while at the highest resolution, the media content is more specifically analyzed, inspected and identified, which thus results in small-scaled media segments.

摘要翻译： 公开了从复合媒体内容以各种分辨率提取语义的一般框架。 具体地，给定媒体流，其可以由包括音频，视觉，文本和图形信息的各种类型的媒体模式组成，所公开的框架描述了如何通过利用和集成不同媒体特征来在不同级别提取各种类型的语义。 该框架的输出是不同尺度的一系列带标签（或注释）的媒体片段。 具体来说，在最低分辨率下，媒体部分的特征更广泛和更广泛，因此它们被更广泛地识别出来。 而在最高分辨率下，更具体地分析，检查和识别媒体内容，从而导致小规模媒体片段。