专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明申请

WO2004042539A9 VIDEO-BASED FACE RECOGNITION USING PROBABILISTIC APPEARANCE MANIFOLDS 审中-公开
标题翻译：基于视觉的面部识别使用概念外观图
公开(公告)号：WO2004042539A9
公开(公告)日：2005-07-21
申请号：PCT/US0335496
申请日：2003-11-06
申请人： HONDA MOTOR CO LTD , YANG MING-HSUAN , HO JEFFREY , LEE KUANG-CHIH
发明人： YANG MING-HSUAN , HO JEFFREY , LEE KUANG-CHIH
IPC分类号： G06F20060101 , G06K9/00 , G06K9/62 , G06K9/64 , G06K9/74 , G06T11/20 , G06T15/00 , H04N7/00 , H04N7/18 , H04N9/47 , H04N13/02 , H04N13/04 , H04N15/00
CPC分类号： G06K9/00335 , G06K9/00275 , G06K9/00288 , G06K9/3208 , G06K9/6232
摘要： The present invention meets these needs by providing temporal coherency to recognition systems. One embodiment of the present invention comprises a manifold recognition module to use a sequence of images for recognition. A manifold training module receives a plurality of training image sequences (e.g. from a video camera), each training image sequence including an individual in a plurality of poses, and establishes relationships between the images of a training image sequence. A probabilistic identity module receives a sequence of recognition images including a target individual for recognition, and identifies the target individual based on the relationship of training images corresponding to the recognition images. An occlusion module masks occluded portions of an individual's face to prevent distorted identifications.
摘要翻译：本发明通过向识别系统提供时间一致性来满足这些需求。本发明的一个实施例包括使用一系列图像进行识别的歧管识别模块。歧管训练模块接收多个训练图像序列（例如，从摄像机），每个训练图像序列包括多个姿势中的个体，并且建立训练图像序列的图像之间的关系。概率识别模块接收包括用于识别的目标个体的识别图像序列，并且基于与识别图像相对应的训练图像的关系来识别目标个体。闭塞模块遮挡个人脸部的遮挡部分以防止变形的识别。

2. 发明申请

WO2004042539A2 VIDEO-BASED FACE RECOGNITION USING PROBABILISTIC APPEARANCE MANIFOLDS 审中-公开
标题翻译：基于视频的人脸识别使用概率外观人工神经网络
公开(公告)号：WO2004042539A2
公开(公告)日：2004-05-21
申请号：PCT/US0335496
申请日：2003-11-06
申请人： HONDA MOTOR CO LTD , YANG MING-HSUAN , HO JEFFREY
发明人： YANG MING-HSUAN , HO JEFFREY , LEE KUANG-CHIH
IPC分类号： G06F20060101 , G06K9/00 , G06K9/62 , G06K9/64 , G06K9/74 , G06T11/20 , G06T15/00 , H04N7/00 , H04N7/18 , H04N9/47 , H04N13/02 , H04N13/04 , H04N15/00 , G06F
CPC分类号： G06K9/00335 , G06K9/00275 , G06K9/00288 , G06K9/3208 , G06K9/6232
摘要： The present invention meets these needs by providing temporal coherency to recognition systems. One embodiment of the present invention comprises a manifold recognition module to use a sequence of images for recognition. A manifold training module receives a plurality of training image sequences (e.g. from a video camera), each training image sequence including an individual in a plurality of poses, and establishes relationships between the images of a training image sequence. A probabilistic identity module receives a sequence of recognition images including a target individual for recognition, and identifies the target individual based on the relationship of training images corresponding to the recognition images. An occlusion module masks occluded portions of an individual's face to prevent distorted identifications.
摘要翻译：本发明通过向识别系统提供时间一致性来满足这些需求。本发明的一个实施例包括使用一系列图像进行识别的歧管识别模块。流形训练模块接收多个训练图像序列（例如来自摄像机），每个训练图像序列包括多个姿势中的个体，并且建立训练图像序列的图像之间的关系。概率识别模块接收包括用于识别的目标个体的识别图像序列，并且基于与识别图像对应的训练图像的关系来识别目标个体。遮挡模块遮挡个人脸部的遮挡部分以防止扭曲的标识。

3. 发明申请

WO2005050533A3 IMAGE CLUSTERING WITH METRIC, LOCAL LINEAR STRUCTURE, AND AFFINE SYMMETRY 审中-公开
标题翻译：具有公制，局部线性结构和AFFINE SYMETRY的图像聚类
公开(公告)号：WO2005050533A3
公开(公告)日：2006-11-23
申请号：PCT/US2004038347
申请日：2004-11-15
申请人： HONDA MOTOR CO LTD , HO JEFFREY , LIM JONGWOO , YANG MING-HSUAN
发明人： HO JEFFREY , LIM JONGWOO , YANG MING-HSUAN
IPC分类号： G06K9/62 , G06K20060101 , G06K9/00 , G06K9/46
CPC分类号： G06K9/6224 , G06K9/6215 , G06K9/6252
摘要： A system and a method are disclosed for clustering (370) images of objects seen from different viewpoints. That is, given an unlabelled set of images of n objects (310), an unsupervised algorithm groups the images into N disjoint subsets such that each subset only contains images of a single object. The clustering method makes use of a broad geometric framework that exploits the interplay between the geometry of appearance manifolds and the symmetry of the 2D affine group.
摘要翻译：公开了一种系统和方法，用于从不同视点看到的对象的聚类（370）图像。也就是说，给定n个对象（310）的未标记的图像集合，无监督的算法将图像分组为N个不相交的子集，使得每个子集仅包含单个对象的图像。聚类方法利用了广泛的几何框架，利用了外观多样性的几何和2D仿射组的对称性之间的相互作用。

4. 发明申请

WO2004044823A3 CLUSTERING APPEARANCES OF OBJECTS UNDER VARYING ILLUMINATION CONDITIONS 审中-公开
标题翻译：在不同的照明条件下对象的聚集外观
公开(公告)号：WO2004044823A3
公开(公告)日：2004-07-15
申请号：PCT/US0335554
申请日：2003-11-06
申请人： HONDA MOTOR CO LTD , YANG MING-HSUAN , HO JEFFREY
发明人： YANG MING-HSUAN , HO JEFFREY
IPC分类号： G06K20060101 , G06K9/00 , G06K9/62
CPC分类号： G06K9/4661 , G06K9/00275 , G06K9/6218
摘要： [0060] Taking a set of unlabeled images of a collection of objects acquired under différent imaging conditions, and decomposing the set into disjoint subsets corresponding to individual objects requires clustering. Appearance-based methods for clustering a set of images (101c) of 3-D objects acquired under varying illumination conditions (100a) can be based on the concept of illumination cones. A clustering problem is equivalent to fmding convex polyhedral eones (301c) in the high-dimensional image space. To efficiently determine the conic structures hidden in the image data, the concept of conic affinity can be used which measures the likelihood of a pair of images belonging to the saine underlying polyhedral cone. Other algorithme can be based on affinity measure based on image gradient comparisons operating directly on die image gradients by comparing the magnitudes and orientations of the image gradient.
摘要翻译：采用在不同成像条件下获取的对象集合的一组未标记图像，并且将该集合分解成对应于各个对象的不相交子集需要聚类。用于聚类在变化的照明条件（100a）下获取的3-D物体的一组图像（101c）的基于外观的方法可以基于照明锥的概念。聚类问题相当于高维图像空间中的多边形凸面（301c）。为了有效地确定隐藏在图像数据中的圆锥形结构，可以使用锥形亲和度的概念，其测量属于下面的多面体锥体中的一对图像的可能性。其他算法可以基于通过比较图像梯度的幅度和方向基于直接在管芯图像梯度上操作的图像梯度比较的亲和度测量。

5. 发明专利

FR3103239A3 Structure de bande flexible capable d’ajuster sa bouche de fixation 未知
公开(公告)号：FR3103239A3
公开(公告)日：2021-05-21
申请号：FR1912862
申请日：2019-11-19
申请人： TSAI CHUNG SHUN , HO JEFFREY
发明人： TSAI CHUNG-SHUN , HO JEFFREY
IPC分类号： F16G11/14 , A45D8/36
摘要： TITRE : Structure de bande flexible capable d’ajuster sa bouche de fixation Une structure de bande flexible capable d'ajuster sa bouche de fixation contient : un corps comprenant deux extrémités insérées dans un orifice traversant d'une bille de positionnement de manière à former une bouche de fixation. Le corps est un groupe tressé de multiples cordes flexibles et multiples arn s, et le corps est flexible et extensible et comprend plusieurs des sections de montage et de multiples parties d’engagement sphériques séparées de la plusieurs sections de montage, respectivement. Le diamètre d’une section de raccord respective est inférieur au diamètre de l’orifice traversant du positionnement balle, un diamètre d'une partie d'engagement respective est deux fois plus que le diamètre de l'orifice traversant du positionnement balle, et la partie d'engagement respective est sphérique. Figure 1

6. 发明专利

FR3103239B3 Structure de bande flexible capable d’ajuster sa bouche de fixation 未知
公开(公告)号：FR3103239B3
公开(公告)日：2021-11-19
申请号：FR1912862
申请日：2019-11-19
申请人： TSAI CHUNG SHUN , HO JEFFREY
发明人： TSAI CHUNG-SHUN , HO JEFFREY
IPC分类号： A45D8/36 , F16G11/14
摘要： TITRE : Structure de bande flexible capable d’ajuster sa bouche de fixation Une structure de bande flexible capable d'ajuster sa bouche de fixation contient : un corps comprenant deux extrémités insérées dans un orifice traversant d'une bille de positionnement de manière à former une bouche de fixation. Le corps est un groupe tressé de multiples cordes flexibles et multiples arn s, et le corps est flexible et extensible et comprend plusieurs des sections de montage et de multiples parties d’engagement sphériques séparées de la plusieurs sections de montage, respectivement. Le diamètre d’une section de raccord respective est inférieur au diamètre de l’orifice traversant du positionnement balle, un diamètre d'une partie d'engagement respective est deux fois plus que le diamètre de l'orifice traversant du positionnement balle, et la partie d'engagement respective est sphérique. Figure 1

7. 发明专利

AU2003301795A1 VIDEO-BASED FACE RECOGNITION USING PROBABILISTIC APPEARANCE MANIFOLDS 未知
公开(公告)号：AU2003301795A1
公开(公告)日：2004-06-07
申请号：AU2003301795
申请日：2003-11-06
申请人： HONDA MOTOR CO LTD
发明人： LEE KUANG-CHIH , YANG MING-HSUAN , HO JEFFREY
IPC分类号： G06F20060101 , G06K9/00 , G06K9/62 , G06K9/64 , G06K9/74 , G06T11/20 , G06T15/00 , H04N7/00 , H04N7/18 , H04N9/47 , H04N13/02 , H04N13/04 , H04N15/00
摘要： The present invention meets these needs by providing temporal coherency to recognition systems. One embodiment of the present invention comprises a manifold recognition module to use a sequence of images for recognition. A manifold training module receives a plurality of training image sequences (e.g. from a video camera), each training image sequence including an individual in a plurality of poses, and establishes relationships between the images of a training image sequence. A probabilistic identity module receives a sequence of recognition images including a target individual for recognition, and identifies the target individual based on the relationship of training images corresponding to the recognition images. An occlusion module masks occluded portions of an individual's face to prevent distorted identifications.

8. 发明专利

AU2003296934A1 CLUSTERING APPEARANCES OF OBJECTS UNDER VARYING ILLUMINATION CONDITIONS 未知
公开(公告)号：AU2003296934A1
公开(公告)日：2004-06-03
申请号：AU2003296934
申请日：2003-11-06
申请人： HONDA MOTOR CO LTD
发明人： YANG MING-HSUAN , HO JEFFREY
IPC分类号： G06K20060101 , G06K9/00 , G06K9/62
摘要： Taking a set of unlabeled images of a collection of objects acquired under different imaging conditions, and decomposing the set into disjoint subsets corresponding to individual objects requires clustering. Appearance-based methods for clustering a set of images of 3-D objects acquired under varying illumination conditions can be based on the concept of illumination cones. A clustering problem is equivalent to finding convex polyhedral cones in the high-dimensional image space. To efficiently determine the conic structures hidden in the image data, the concept of conic affinity can be used which measures the likelihood of a pair of images belonging to the same underlying polyhedral cone. Other algorithms can be based on affinity measure based on image gradient comparisons operating directly on the image gradients by comparing the magnitudes and orientations of the image gradient.

9. 发明专利

AU2003296934A8 Clustering appearances of objects under varying illumination conditions 未知
公开(公告)号：AU2003296934A8
公开(公告)日：2004-06-03
申请号：AU2003296934
申请日：2003-11-06
申请人： HONDA MOTOR CO LTD
发明人： YANG MING-HSUAN , HO JEFFREY
IPC分类号： G06K20060101 , G06K9/00 , G06K9/62
摘要： Taking a set of unlabeled images of a collection of objects acquired under different imaging conditions, and decomposing the set into disjoint subsets corresponding to individual objects requires clustering. Appearance-based methods for clustering a set of images of 3-D objects acquired under varying illumination conditions can be based on the concept of illumination cones. A clustering problem is equivalent to finding convex polyhedral cones in the high-dimensional image space. To efficiently determine the conic structures hidden in the image data, the concept of conic affinity can be used which measures the likelihood of a pair of images belonging to the same underlying polyhedral cone. Other algorithms can be based on affinity measure based on image gradient comparisons operating directly on the image gradients by comparing the magnitudes and orientations of the image gradient.

10. 发明专利

DE60326391D1 未知
公开(公告)号：DE60326391D1
公开(公告)日：2009-04-09
申请号：DE60326391
申请日：2003-11-06
申请人： HONDA MOTOR CO LTD
发明人： YANG MING-HSUAN , HO JEFFREY
IPC分类号： G06K9/62 , G06K20060101 , G06K9/00
摘要： Taking a set of unlabeled images of a collection of objects acquired under different imaging conditions, and decomposing the set into disjoint subsets corresponding to individual objects requires clustering. Appearance-based methods for clustering a set of images of 3-D objects acquired under varying illumination conditions can be based on the concept of illumination cones. A clustering problem is equivalent to finding convex polyhedral cones in the high-dimensional image space. To efficiently determine the conic structures hidden in the image data, the concept of conic affinity can be used which measures the likelihood of a pair of images belonging to the same underlying polyhedral cone. Other algorithms can be based on affinity measure based on image gradient comparisons operating directly on the image gradients by comparing the magnitudes and orientations of the image gradient.

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式