-
公开(公告)号：US08027541B2
公开(公告)日：2011-09-27
申请号：US11725129
申请日：2007-03-15
申请人： Gang Hua , Steven M. Drucker , Michael Revow , Paul A. Viola , Richard Zemel
发明人： Gang Hua , Steven M. Drucker , Michael Revow , Paul A. Viola , Richard Zemel
IPC分类号： G06K9/46
CPC分类号： G06K9/00228 , G06K9/6251
摘要： A system for organizing images includes an extraction component that extracts visual information (e.g., faces, scenes, etc.) from the images. The extracted visual information is provided to a comparison component which computes similarity confidence data between the extracted visual information. The similarity confidence data is an indication of the likelihood that items of extracted visual information are similar. The comparison component then generates a visual distribution of the extracted visual information based upon the similarity confidence data. The visual distribution can include groupings of the extracted visual information based on computed similarity confidence data. For example, the visual distribution can be a two-dimensional layout of faces organized based on the computed similarity confidence data—with faces in closer proximity faces computed to have a greater probability of representing the same person. The visual distribution can then be utilized by a user to sort, organize and/or tag images.
摘要翻译： 用于组织图像的系统包括从图像中提取视觉信息（例如，面部，场景等）的提取组件。 提取的视觉信息被提供给计算提取的视觉信息之间的相似性置信度数据的比较部件。 相似性置信度数据是提取的视觉信息的项目相似的可能性的指示。 然后，比较组件基于相似性置信度数据生成所提取的视觉信息的视觉分布。 视觉分布可以包括基于计算的相似性置信度数据提取的视觉信息的分组。 例如，视觉分布可以是基于所计算的相似性置信度数据组织的面部的二维布局，其中更接近的面中的面被计算为具有更大的代表同一人的概率。 然后用户可以利用视觉分布来对图像进行分类，组织和/或标记。
-
公开(公告)号：US20080226174A1
公开(公告)日：2008-09-18
申请号：US11725129
申请日：2007-03-15
申请人： Gang Hua , Steven M. Drucker , Michael Revow , Paul A. Viola , Richard Zemel
发明人： Gang Hua , Steven M. Drucker , Michael Revow , Paul A. Viola , Richard Zemel
CPC分类号： G06K9/00228 , G06K9/6251
摘要： A system for organizing images includes an extraction component that extracts visual information (e.g., faces, scenes, etc.) from the images. The extracted visual information is provided to a comparison component which computes similarity confidence data between the extracted visual information. The similarity confidence data is an indication of the likelihood that items of extracted visual information are similar. The comparison component then generates a visual distribution of the extracted visual information based upon the similarity confidence data. The visual distribution can include groupings of the extracted visual information based on computed similarity confidence data. For example, the visual distribution can be a two-dimensional layout of faces organized based on the computed similarity confidence data—with faces in closer proximity faces computed to have a greater probability of representing the same person. The visual distribution can then be utilized by a user to sort, organize and/or tag images.
摘要翻译： 用于组织图像的系统包括从图像中提取视觉信息（例如，面部，场景等）的提取组件。 提取的视觉信息被提供给计算提取的视觉信息之间的相似性置信度数据的比较部件。 相似性置信度数据是提取的视觉信息的项目相似的可能性的指示。 然后，比较组件基于相似性置信度数据生成所提取的视觉信息的视觉分布。 视觉分布可以包括基于计算的相似性置信度数据提取的视觉信息的分组。 例如，视觉分布可以是基于所计算的相似性置信度数据组织的面部的二维布局，其中更接近的面中的面被计算为具有更大的代表同一人的概率。 然后用户可以利用视觉分布来对图像进行分类，组织和/或标记。
-
3.发明授权公开(公告)号：US07936906B2
公开(公告)日：2011-05-03
申请号：US11763909
申请日：2007-06-15
申请人： Gang Hua , Paul A Viola , Steven M. Drucker , Michael Revow
发明人： Gang Hua , Paul A Viola , Steven M. Drucker , Michael Revow
IPC分类号： G06K9/00
CPC分类号： G06K9/00288 , G06K9/6232
摘要： Systems and methods are described for face recognition using discriminatively trained orthogonal rank one tensor projections. In an exemplary system, images are treated as tensors, rather than as conventional vectors of pixels. During runtime, the system designs visual features—embodied as tensor projections—that minimize intraclass differences between instances of the same face while maximizing interclass differences between the face and faces of different people. Tensor projections are pursued sequentially over a training set of images and take the form of a rank one tensor, i.e., the outer product of a set of vectors. An exemplary technique ensures that the tensor projections are orthogonal to one another, thereby increasing ability to generalize and discriminate image features over conventional techniques. Orthogonality among tensor projections is maintained by iteratively solving an ortho-constrained eigenvalue problem in one dimension of a tensor while solving unconstrained eigenvalue problems in additional dimensions of the tensor.
摘要翻译： 使用区分训练的正交秩一张量投影描述用于人脸识别的系统和方法。 在示例性系统中，图像被视为张量，而不是像传统的像素矢量。 在运行期间，系统设计视觉特征 - 体现为张量投影 - 最大限度地减少不同人脸部和脸部之间的类间差异，从而最大限度地减少同一脸部实例之间的差异。 张量投影在训练图像集上顺序追溯，并采取一级张量的形式，即一组向量的外积。 示例性技术确保张量投影彼此正交，从而增加了与常规技术相比的概括和区分图像特征的能力。 通过迭代求解张量的一维中的邻域约束特征值问题，同时解决张量的附加维度中的无约束特征值问题，维持张量投影中的正交性。
-
4.发明申请公开(公告)号：US20080310687A1
公开(公告)日：2008-12-18
申请号：US11763909
申请日：2007-06-15
申请人： Gang Hua , Paul A. Viola , Steven M. Drucker , Michael Revow
发明人： Gang Hua , Paul A. Viola , Steven M. Drucker , Michael Revow
IPC分类号： G06K9/00
CPC分类号： G06K9/00288 , G06K9/6232
摘要： Systems and methods are described for face recognition using discriminatively trained orthogonal rank one tensor projections. In an exemplary system, images are treated as tensors, rather than as conventional vectors of pixels. During runtime, the system designs visual features—embodied as tensor projections—that minimize intraclass differences between instances of the same face while maximizing interclass differences between the face and faces of different people. Tensor projections are pursued sequentially over a training set of images and take the form of a rank one tensor, i.e., the outer product of a set of vectors. An exemplary technique ensures that the tensor projections are orthogonal to one another, thereby increasing ability to generalize and discriminate image features over conventional techniques. Orthogonality among tensor projections is maintained by iteratively solving an ortho-constrained eigenvalue problem in one dimension of a tensor while solving unconstrained eigenvalue problems in additional dimensions of the tensor.
摘要翻译： 使用区分训练的正交秩一张量投影描述用于人脸识别的系统和方法。 在示例性系统中，图像被视为张量，而不是像传统的像素矢量。 在运行期间，系统设计视觉特征 - 体现为张量投影 - 最大限度地减少不同人脸部和脸部之间的类间差异，从而最大限度地减少同一脸部实例之间的差异。 张量投影在训练图像集上顺序追溯，并采取一级张量的形式，即一组向量的外积。 示例性技术确保张量投影彼此正交，从而增加了与常规技术相比的概括和区分图像特征的能力。 通过迭代求解张量的一维中的邻域约束特征值问题，同时解决张量的附加维度中的无约束特征值问题，维持张量投影中的正交性。
-
公开(公告)号：US09372873B2
公开(公告)日：2016-06-21
申请号：US13616617
申请日：2012-09-14
CPC分类号： G06F17/3053 , G06F17/30126 , G06F17/30274 , G06F17/30554 , G06F17/30713 , G06F17/30867 , G06F17/30905
摘要： Methods are provided for displaying image results responsive to a search query. In addition to displaying responsive results for a query, responsive results are also provided for related queries. The results are ordered along a plurality of display axes, including at least one axis corresponding to the ordering of the various search queries. The results can be displayed in an aligned or non-aligned manner. The results can then be translated along one or more of the display axes to allow a user to browse the various results.
-
公开(公告)号：US20130194294A1
公开(公告)日：2013-08-01
申请号：US13359508
申请日：2012-01-27
申请人： Jian Zhao , Steven M. Drucker , Danyel Fisher , Donald Brinkman
发明人： Jian Zhao , Steven M. Drucker , Danyel Fisher , Donald Brinkman
IPC分类号： G09G5/00
CPC分类号： G06F17/30389
摘要： Technology for rendering representations of multi-faceted data are disclosed. As one example, the technology includes organizing and rendering multiple subsets of a dataset according to temporal or other linear attributes, e.g., for visual comparison and/or other visualization. A collection of subsets may be determined in response to a selection of a facet having multiple facet attributes. Each subset may include the entries of the dataset that have the facet attribute corresponding to that subset. Optionally, the multiple subsets may be rendered in alignment with corresponding portions of a hierarchical depiction, e.g., to visually represent the parameters of the subsets.
摘要翻译： 公开了表示多方面数据的技术。 作为一个示例，该技术包括根据时间或其他线性属性来组织和呈现数据集的多个子集，例如用于视觉比较和/或其他可视化。 可以响应于具有多个方面属性的方面的选择来确定子集合。 每个子集可以包括具有与该子集对应的facet属性的数据集的条目。 可选地，多个子集可以被渲染成与分层描述的相应部分对齐，例如，以视觉上表示子集的参数。
-
公开(公告)号：US08205172B2
公开(公告)日：2012-06-19
申请号：US11097720
申请日：2005-03-31
IPC分类号： G06F3/048
CPC分类号： G06F17/30884
摘要： The present subject matter describes a graphical web browser toolbar the stores visual representations of user-selected web pages. A saved page is included in the toolbar as a tile that shows a thumbnail image of at least a portion of the page. A user may customize the thumbnail image by zooming onto an area of the page. The user selects a tile from the toolbar to show an enlarged version of the thumbnail image with data from the web page, such as a link to the web page. A comments section allows a user to enter comments related to the web page into the tile. Tiles may be arranged according to groups as configured by the user. The toolbar may be stored and transmitted to a remote location. Using the described techniques, a web shopper can save images in a graphical toolbar and send items saved therein to an acquaintance.
摘要翻译： 本主题描述了图形网络浏览器工具栏，其存储用户选择的网页的视觉表示。 保存的页面被包括在工具栏中，作为显示页面的至少一部分的缩略图的瓦片。 用户可以通过缩放到页面的区域来自定义缩略图图像。 用户从工具栏中选择一个瓦片，以显示具有来自网页的数据的缩略图像的放大版本，例如到网页的链接。 评论部分允许用户将有关网页的评论输入到图块中。 可以根据用户配置的组来布置瓦片。 工具栏可以存储并发送到远程位置。 使用所描述的技术，网络购物者可以将图像保存在图形工具栏中，并将保存在其中的项目发送给熟人。
-
公开(公告)号：US07769832B2
公开(公告)日：2010-08-03
申请号：US11539963
申请日：2006-10-10
IPC分类号： G06F15/16
CPC分类号： G06F17/30029 , G06F17/30058 , G06F17/30873
摘要： A system and method facilitating media browsing is provided. The invention includes a browsing component and a media display system that provides a focus media and related media. The focus media and the related media are associated based, at least in part, upon metadata. The invention further provides for an association identifier that identifies related media based, at least in part, upon metadata associated with the focus media and the related media. The association identifier can further cluster the related media based upon metadata associated with the related media.
摘要翻译： 提供了一种促进媒体浏览的系统和方法。 本发明包括提供聚焦媒体和相关媒体的浏览组件和媒体显示系统。 焦点媒体和相关媒体至少部分地基于元数据。 本发明还提供了至少部分地基于与聚焦媒体和相关媒体相关联的元数据来识别相关媒体的关联标识符。 关联标识符可以基于与相关媒体相关联的元数据进一步聚集相关媒体。
-
公开(公告)号：US07725829B1
公开(公告)日：2010-05-25
申请号：US11009115
申请日：2004-12-10
IPC分类号： G06F3/00
CPC分类号： G11B27/105 , G06F17/30053
摘要： A system and method for facilitating non-linear viewing of media is provided. The system facilitates non-linear viewing of media by providing a scene selector that scans a digitized media and selects a scene in the digitized media and a metadata generator that produces metadata associated with the scenes and relates the metadata to the selected scene. With the scenes annotated with metadata, a playlist generator can generate a playlist of related scenes based on user inputs like queries and a playlist updater can adapt the playlist based on user reaction to the displayed scenes. The scenes can be displayed on a variety of devices exhibiting various levels of intelligence. The displays can be distributed as can the system.
-
公开(公告)号：US07519907B2
公开(公告)日：2009-04-14
申请号：US10633776
申请日：2003-08-04
IPC分类号： G06F3/00
CPC分类号： G06T15/503 , G06T11/60
摘要： A system and method for editing images. A simple but powerful image stack is employed in creating an enhanced image from a stack of registered images. This paradigm combines pixels using multi-image operations on the image stack. Image Stacks can help create group photographs, create high dynamic range images, combine images captured under different lighting conditions, remove unwanted objects from images, and combine images captured at different times and with different focal lengths.
摘要翻译： 一种用于编辑图像的系统和方法。 一个简单但功能强大的图像堆栈用于从一堆已注册图像创建增强图像。 该范例使用图像堆叠上的多图像操作组合像素。 图像堆叠可以帮助创建组合照片，创建高动态范围图像，组合在不同照明条件下捕获的图像，从图像中删除不需要的对象，并组合在不同时间和不同焦距捕获的图像。
-
-
-
-
-
-
-
-
-
搜索结果
80 条记录 (耗时 0.036 秒)
