公开(公告)号：US08781175B2

公开(公告)日：2014-07-15

申请号：US13493564

申请日：2012-06-11

申请人： James Z. Wang ,  Jia Li ,  Lei Yao ,  Poonam Suryanarayan ,  Mu Qiao

发明人： James Z. Wang ,  Jia Li ,  Lei Yao ,  Poonam Suryanarayan ,  Mu Qiao

IPC分类号： G06K9/00

CPC分类号： G06K9/6267 ,  G06K9/00624 ,  G06K9/46 ,  G06K9/4652 ,  G06K9/4671 ,  G06K9/6228 ,  H04N5/23222

摘要： A comprehensive system to enhance the aesthetic quality of the photographs captured by mobile consumers provides on-site composition and aesthetics feedback through retrieved examples. Composition feedback is qualitative in nature and responds by retrieving highly aesthetic exemplar images from the corpus which are similar in content and composition to the snapshot. Color combination feedback provides confidence on the snapshot to contain good color combinations. Overall aesthetics feedback predicts the aesthetic ratings for both color and monochromatic images. An algorithm is used to provide ratings for color images, while new features and a new model are developed to treat monochromatic images. This system was designed keeping the next generation photography needs in mind and is the first of its kind. The feedback rendered is guiding and intuitive in nature. It is computed in situ while requiring minimal input from the user.

公开(公告)号：US20090083332A1

公开(公告)日：2009-03-26

申请号：US12234159

申请日：2008-09-19

申请人： Ritendra Datta ,  Dhiraj Joshi ,  Jia Li ,  James Z. Wang

发明人： Ritendra Datta ,  Dhiraj Joshi ,  Jia Li ,  James Z. Wang

IPC分类号： G06F17/30

CPC分类号： G06K9/6263 ,  G06F16/58

摘要： A principled, probabilistic approach to meta-learning acts as a go-between for a ‘black-box’ image annotation system and its users. Inspired by inductive transfer, the approach harnesses available information, including the black-box model's performance, the image representations, and a semantic lexicon ontology. Being computationally ‘lightweight.’ the meta-learner efficiently re-trains over time, to improve and/or adapt to changes. The black-box annotation model is not required to be re-trained, allowing computationally intensive algorithms to be used. Both batch and online annotation settings are accommodated. A “tagging over time” approach produces progressively better annotation, significantly outperforming the black-box as well as the static form of the meta-learner, on real-world data.

摘要翻译： 元学习的原则性，概率性方法作为“黑盒子”图像注释系统及其用户的一个中介。 灵感来自感性传递，该方法利用了可用的信息，包括黑箱模型的性能，图像表示和语义词典本体。 在计算上“轻量级”。 元学习者随着时间的推移有效地重新训练，以改善和/或适应变化。 黑箱注释模型不需要重新训练，允许使用计算密集型算法。 批量和在线注释设置都可以收录。 随着时间的推移，“标记”方法可以逐渐更好的注释，显着优于实体数据的黑盒子和元学习者的静态形式。

公开(公告)号：US20080285860A1

公开(公告)日：2008-11-20

申请号：US12116578

申请日：2008-05-07

申请人： Ritendra Datta ,  Jia Li ,  James Z. Wang

发明人： Ritendra Datta ,  Jia Li ,  James Z. Wang

IPC分类号： G06K9/62 ,  G06K9/00

CPC分类号： G06K9/6228 ,  G06K9/00624 ,  G06K9/4652 ,  G06K9/4671

摘要： The aesthetic quality of a picture is automatically inferred using visual content as a machine learning problem using, for example, a peer-rated, on-line photo sharing Website as data source. Certain visual features of images are extracted based on the intuition that they can discriminate between aesthetically pleasing and displeasing images. A one-dimensional support vector machine is used to identify features that have noticeable correlation with the community-based aesthetics ratings. Automated classifiers are constructed using the support vector machines and classification trees, with a simple feature selection heuristic being applied to eliminate irrelevant features. Linear regression on polynomial terms of the features is also applied to infer numerical aesthetics ratings.

摘要翻译： 使用例如同行评级的在线照片共享网站作为数据源，使用视觉内容作为机器学习问题自动推断图片的美学品质。 基于它们可以在美学上令人不愉快的图像之间区分的直觉来提取图像的某些视觉特征。 一维支持向量机用于识别与基于社区的美学评级具有显着相关性的特征。 使用支持向量机和分类树构建自动分类器，并采用简单的特征选择启发式来消除不相关的特征。 特征的多项式项的线性回归也用于推断数字美学评级。

公开(公告)号：US07941009B2

公开(公告)日：2011-05-10

申请号：US11872260

申请日：2007-10-15

申请人： Jia Li ,  James Z. Wang

发明人： Jia Li ,  James Z. Wang

IPC分类号： G06K9/54 ,  G06F7/00

CPC分类号： G06K9/00664 ,  G06F17/3025 ,  G06F17/30262 ,  G06F17/30265 ,  G06K9/6215 ,  G06K9/6226

摘要： A computerized annotation method achieves real-time operation and better optimization properties while preserving the architectural advantages of the generative modeling approach. A novel clustering algorithm for objects is represented by discrete distributions, or bags of weighted vectors, thereby minimizing the total within cluster distance, a criterion used by the k-means algorithm. A new mixture modeling method, the hypothetical local mapping (HLM) method, is used to efficiently build a probability measure on the space of discrete distributions. Thus, in accord with the invention every image is characterized by a statistical distribution. The profiling model specifies a probability law for distributions directly.

摘要翻译： 计算机化注释方法实现了实时操作和更好的优化属性，同时保留了生成建模方法的架构优势。 用于对象的新颖的聚类算法由离散分布或包的加权向量表示，从而使簇距离内的总数最小化，这是k-means算法使用的标准。 一种新的混合建模方法，即假设局部映射（HLM）方法，用于有效构建离散分布空间的概率测度。 因此，根据本发明，每个图像的特征在于统计分布。 分析模型直接指定分布概率定律。

公开(公告)号：US07929805B2

公开(公告)日：2011-04-19

申请号：US11668853

申请日：2007-01-30

申请人： James Z. Wang ,  Ritendra Datta ,  Jia Li

发明人： James Z. Wang ,  Ritendra Datta ,  Jia Li

IPC分类号： G06K9/32

CPC分类号： G06K9/00

摘要： In a system and method for the generation of attack-resistant, user-friendly, image-based CAPTCHAs (Completely Automated Public test to Tell Computers and Humans Apart), controlled distortions are applied to randomly chosen images and presented to a user for annotation from a given list of words. An image is presented that contains multiple connected but independent images with the borders between them distorted or otherwise visually obfuscated in a way that a computer cannot distinguish the borders and a user selects near the center of one of the images. The distortions are performed in a way that satisfies the incongruous requirements of low perceptual degradation and high resistance to attack by content-based image retrieval systems. Word choices are carefully generated to avoid ambiguity as well as to avoid attacks based on the choices themselves.

摘要翻译： 在用于生成抗攻击，用户友好的基于图像的CAPTCHAs（完全自动公共测试以告知计算机和人类）的系统和方法中，控制的失真被应用于随机选择的图像，并呈现给用户以从 给定的单词列表。 呈现包含多个连接但独立的图像的图像，其中它们之间的边界被扭曲或者以计算机无法区分边界并且用户在图像之一附近的中心附近选择的视觉模糊处理。 这种扭曲是以满足基于内容的图像检索系统的低感知降解和高抵抗攻击的不协调要求的方式进行的。 仔细地生成词选择以避免歧义，并避免基于选择本身的攻击。

公开(公告)号：US08755596B2

公开(公告)日：2014-06-17

申请号：US13542326

申请日：2012-07-05

申请人： Ritendra Datta ,  Jia Li ,  James Z. Wang

发明人： Ritendra Datta ,  Jia Li ,  James Z. Wang

IPC分类号： G06K9/62

CPC分类号： G06K9/6228 ,  G06K9/00624 ,  G06K9/4652 ,  G06K9/4671

摘要： The aesthetic quality of a picture is automatically inferred using visual content as a machine learning problem using, for example, a peer-rated, on-line photo sharing Website as data source. Certain visual features of images are extracted based on the intuition that they can discriminate between aesthetically pleasing and displeasing images. A one-dimensional support vector machine is used to identify features that have noticeable correlation with the community-based aesthetics ratings. Automated classifiers are constructed using the support vector machines and classification trees, with a simple feature selection heuristic being applied to eliminate irrelevant features. Linear regression on polynomial terms of the features is also applied to infer numerical aesthetics ratings.

摘要翻译： 使用例如同行评级的在线照片共享网站作为数据源，使用视觉内容作为机器学习问题自动推断图片的美学品质。 基于它们可以在美学上令人不愉快的图像之间区分的直觉来提取图像的某些视觉特征。 一维支持向量机用于识别与基于社区的美学评级具有显着相关性的特征。 使用支持向量机和分类树构建自动分类器，并采用简单的特征选择启发式来消除不相关的特征。 特征的多项式项的线性回归也用于推断数字美学评级。

公开(公告)号：US20090204637A1

公开(公告)日：2009-08-13

申请号：US11872260

申请日：2007-10-15

申请人： Jia Li ,  James Z. Wang

发明人： Jia Li ,  James Z. Wang

IPC分类号： G06F17/00

CPC分类号： G06K9/00664 ,  G06F17/3025 ,  G06F17/30262 ,  G06F17/30265 ,  G06K9/6215 ,  G06K9/6226

摘要： A computerized annotation method achieves real-time operation and better optimization properties while preserving the architectural advantages of the generative modeling approach. A novel clustering algorithm for objects is represented by discrete distributions, or bags of weighted vectors, thereby minimizing the total within cluster distance, a criterion used by the k-means algorithm. A new mixture modeling method, the hypothetical local mapping (HLM) method, is used to efficiently build a probability measure on the space of discrete distributions. Thus, in accord with the invention every image is characterized by a statistical distribution. The profiling model specifies a probability law for distributions directly.

摘要翻译： 计算机化注释方法实现了实时操作和更好的优化属性，同时保留了生成建模方法的架构优势。 用于对象的新颖的聚类算法由离散分布或包的加权向量表示，从而使簇距离内的总数最小化，这是k-means算法使用的标准。 一种新的混合建模方法，即假设局部映射（HLM）方法，用于有效构建离散分布空间的概率测度。 因此，根据本发明，每个图像的特征在于统计分布。 分析模型直接指定分布概率定律。

公开(公告)号：US20130011070A1

公开(公告)日：2013-01-10

申请号：US13542326

申请日：2012-07-05

申请人： Ritendra Datta ,  Jia Li ,  James Z. Wang

发明人： Ritendra Datta ,  Jia Li ,  James Z. Wang

IPC分类号： G06K9/62 ,  G06K9/46

CPC分类号： G06K9/6228 ,  G06K9/00624 ,  G06K9/4652 ,  G06K9/4671

摘要： The aesthetic quality of a picture is automatically inferred using visual content as a machine learning problem using, for example, a peer-rated, on-line photo sharing Website as data source. Certain visual features of images are extracted based on the intuition that they can discriminate between aesthetically pleasing and displeasing images. A one-dimensional support vector machine is used to identify features that have noticeable correlation with the community-based aesthetics ratings. Automated classifiers are constructed using the support vector machines and classification trees, with a simple feature selection heuristic being applied to eliminate irrelevant features. Linear regression on polynomial terms of the features is also applied to infer numerical aesthetics ratings.

摘要翻译： 使用例如同行评级的在线照片共享网站作为数据源，使用视觉内容作为机器学习问题自动推断图片的美学品质。 基于它们可以在美学上令人不愉快的图像之间区分的直觉来提取图像的某些视觉特征。 一维支持向量机用于识别与基于社区的美学评级具有显着相关性的特征。 使用支持向量机和分类树构建自动分类器，并采用简单的特征选择启发式来消除不相关的特征。 特征的多项式项的线性回归也用于推断数字美学评级。

公开(公告)号：US20120268612A1

公开(公告)日：2012-10-25

申请号：US13493564

申请日：2012-06-11

申请人： James Z. Wang ,  Jia Li ,  Lei Yao ,  Poonam Suryanarayan ,  Mu Qiao

发明人： James Z. Wang ,  Jia Li ,  Lei Yao ,  Poonam Suryanarayan ,  Mu Qiao

IPC分类号： H04N5/225

CPC分类号： G06K9/6267 ,  G06K9/00624 ,  G06K9/46 ,  G06K9/4652 ,  G06K9/4671 ,  G06K9/6228 ,  H04N5/23222

摘要： A comprehensive system to enhance the aesthetic quality of the photographs captured by mobile consumers provides on-site composition and aesthetics feedback through retrieved examples. Composition feedback is qualitative in nature and responds by retrieving highly aesthetic exemplar images from the corpus which are similar in content and composition to the snapshot. Color combination feedback provides confidence on the snapshot to contain good color combinations. Overall aesthetics feedback predicts the aesthetic ratings for both color and monochromatic images. An algorithm is used to provide ratings for color images, while new features and a new model are developed to treat monochromatic images. This system was designed keeping the next generation photography needs in mind and is the first of its kind. The feedback rendered is guiding and intuitive in nature. It is computed in situ while requiring minimal input from the user.

摘要翻译： 提高移动消费者拍摄的照片的美学质量的全面系统通过检索的例子提供现场构图和美学反馈。 组合反馈本质上是定性的，并通过从语料库中检索与内容和构图相似的快照的高度审美的示例图像来进行响应。 颜色组合反馈提供了对快照的信心，以包含良好的颜色组合。 整体美学反馈预测了彩色和单色图像的审美等级。 一种算法用于提供彩色图像的等级，同时开发新特征和新模型来处理单色图像。 该系统的设计保持了下一代摄影的需要，是同类产品中的第一个。 呈现的反馈本质上是指导和直观的。 它是原地计算的，同时需要用户的最小输入。

公开(公告)号：US06633817B1

公开(公告)日：2003-10-14

申请号：US09474929

申请日：1999-12-29

申请人： Michael G. Walker ,  James Z. Wang ,  Eldar Y. Giladi

发明人： Michael G. Walker ,  James Z. Wang ,  Eldar Y. Giladi

IPC分类号： G01N3348

CPC分类号： G06F19/22 ,  G06F19/24

摘要： A method and system for generating and searching a tree-structured index of window vectors that represent database sequences comprise a window vector generation module, a tree-structured index generation module, a query sequence partitioning module, and a retrieval component. The window vector generation module partitions a database sequence into a plurality of overlapping windows. Each window has a fixed length W comprising a fixed number of nucleotides, and the offset among windows is determined by a parameter &Dgr;. The window vector generation module then maps each database sequence window into a window vector. The database sequence window vector indicates the frequency of appearance of each k-tuple in the corresponding database sequence window. The tree-structured index generation module then generates a tree-structured index using the database sequence window vectors. The query sequence partitioning module partitions a query sequence into a plurality of windows and maps each query sequence window into a query sequence window vector. Each query sequence window vector is then compared against the tree-structured index to locate the database sequences that are similar to the query sequence. The list of database sequences that are similar to the query sequence is then returned as the result of the search.

摘要翻译： 用于生成和搜索表示数据库序列的窗口向量的树结构索引的方法和系统包括窗口向量生成模块，树结构索引生成模块，查询序列分区模块和检索组件。 窗口向量生成模块将数据库序列分割成多个重叠窗口。 每个窗口具有包含固定数目的核苷酸的固定长度W，并且窗口之间的偏移由参数Delta确定。 然后，窗口向量生成模块将每个数据库序列窗口映射到窗口向量中。 数据库序列窗口向量表示相应数据库序列窗口中每个k元组的出现频率。 然后，树结构索引生成模块使用数据库序列窗口向量生成树结构索引。 查询序列分区模块将查询序列分割成多个窗口，并将每个查询序列窗口映射到查询序列窗口向量中。 然后将每个查询序列窗口向量与树结构索引进行比较，以定位与查询序列相似的数据库序列。 作为搜索结果，返回与查询序列相似的数据库序列列表。