公开(公告)号：US20140376770A1

公开(公告)日：2014-12-25

申请号：US13926882

申请日：2013-06-25

申请人： David Nister ,  Piotr Dollar ,  Wolf Kienzle ,  Mladen Radojevic ,  Matthew S. Ashman ,  Ivan Stojiljkovic ,  Magdalena Vukosavljevic

发明人： David Nister ,  Piotr Dollar ,  Wolf Kienzle ,  Mladen Radojevic ,  Matthew S. Ashman ,  Ivan Stojiljkovic ,  Magdalena Vukosavljevic

IPC分类号： G06T7/00 ,  G06K9/62 ,  G06K9/00

CPC分类号： G06K9/6256 ,  G06F3/017 ,  G06K9/00375 ,  G06K9/00536 ,  G06K9/6292 ,  G06T7/73

摘要： A method of object detection includes receiving a first image taken by a first stereo camera, receiving a second image taken by a second stereo camera, and offsetting the first image relative to the second image by an offset distance selected such that each corresponding pixel of offset first and second images depict a same object locus if the object locus is at an assumed distance from the first and second stereo cameras. The method further includes locating a target object in the offset first and second images.

摘要翻译： 一种物体检测方法包括：接收由第一立体相机拍摄的第一图像，接收由第二立体相机拍摄的第二图像，以及相对于第二图像偏移所述第一图像的偏移距离，使得每个对应的偏移像素 如果对象轨迹距离第一立体相机和第二立体相机假定距离，则第一和第二图像描绘相同的对象轨迹。 该方法还包括将偏移的第一和第二图像中的目标对象定位。

公开(公告)号：US09881354B2

公开(公告)日：2018-01-30

申请号：US13421307

申请日：2012-03-15

申请人： Johannes Peter Kopf ,  Sing Bing Kang ,  Wolf Kienzle ,  Steven M. Drucker

发明人： Johannes Peter Kopf ,  Sing Bing Kang ,  Wolf Kienzle ,  Steven M. Drucker

IPC分类号： G06K9/00 ,  G06T3/40 ,  G06T5/00

CPC分类号： G06T3/4038 ,  G06T5/005

摘要： Described is a technology by which an image such as a stitched panorama is automatically cropped based upon predicted quality data with respect to filling missing pixels. The image may be completed, including by completing only those missing pixels that remain after cropping. Predicting quality data may be based on using restricted search spaces corresponding to the missing pixels. The crop is computed based upon the quality data, in which the crop is biased towards including original pixels and excluding predicted low quality pixels. Missing pixels are completed by using restricted search spaces to find replacement values for the missing pixels, and may use histogram matching for texture synthesis.

公开(公告)号：US07886266B2

公开(公告)日：2011-02-08

申请号：US11278949

申请日：2006-04-06

申请人： Wolf Kienzle ,  Kumar H. Chellapilla

发明人： Wolf Kienzle ,  Kumar H. Chellapilla

IPC分类号： G06F9/44

CPC分类号： G10L15/07

摘要： The subject disclosure pertains to systems and methods for personalization of a recognizer. In general, recognizers can be used to classify input data. During personalization, a recognizer is provided with samples specific to a user, entity or format to improve performance for the specific user, entity or format. Biased regularization can be utilized during personalization to maintain recognizer performance for non-user specific input. In one aspect, regularization can be biased to the original parameters of the recognizer, such that the recognizer is not modified excessively during personalization.

摘要翻译： 本发明涉及用于识别器个性化的系统和方法。 通常，识别器可用于对输入数据进行分类。 在个性化期间，向识别器提供特定于用户，实体或格式的样本，以提高特定用户，实体或格式的性能。 在个性化过程中可以利用偏置正则化来维持非用户特定输入的识别器性能。 在一个方面，正则化可以偏向识别器的原始参数，使得识别器在个性化期间不被过度修改。

公开(公告)号：US20130243320A1

公开(公告)日：2013-09-19

申请号：US13421307

申请日：2012-03-15

申请人： Johannes Peter Kopf ,  Sing Bing Kang ,  Wolf Kienzle ,  Steven M. Drucker

发明人： Johannes Peter Kopf ,  Sing Bing Kang ,  Wolf Kienzle ,  Steven M. Drucker

IPC分类号： G06K9/36

CPC分类号： G06T3/4038 ,  G06T5/005

摘要： Described is a technology by which an image such as a stitched panorama is automatically cropped based upon predicted quality data with respect to filling missing pixels. The image may be completed, including by completing only those missing pixels that remain after cropping. Predicting quality data may be based on using restricted search spaces corresponding to the missing pixels. The crop is computed based upon the quality data, in which the crop is biased towards including original pixels and excluding predicted low quality pixels. Missing pixels are completed by using restricted search spaces to find replacement values for the missing pixels, and may use histogram matching for texture synthesis.

摘要翻译： 描述了一种基于相对于填充缺失像素的预测质量数据自动裁剪诸如缝合全景图像的技术。 可以完成图像，包括仅完成裁剪后剩余的那些丢失的像素。 预测质量数据可以基于对应于缺失像素的限制搜索空间。 基于质量数据计算作物，其中作物偏向于包括原始像素并排除预测的低质量像素。 通过使用限制搜索空间来找到丢失的像素的替换值，可以完成缺少的像素，并且可以使用纹理合成的直方图匹配。

公开(公告)号：US20070239450A1

公开(公告)日：2007-10-11

申请号：US11278949

申请日：2006-04-06

申请人： Wolf Kienzle ,  Kumar Chellapilla

发明人： Wolf Kienzle ,  Kumar Chellapilla

IPC分类号： G10L15/06

CPC分类号： G10L15/07

摘要： The subject disclosure pertains to systems and methods for personalization of a recognizer. In general, recognizers can be used to classify input data. During personalization, a recognizer is provided with samples specific to a user, entity or format to improve performance for the specific user, entity or format. Biased regularization can be utilized during personalization to maintain recognizer performance for non-user specific input. In one aspect, regularization can be biased to the original parameters of the recognizer, such that the recognizer is not modified excessively during personalization.

摘要翻译： 本发明涉及用于识别器个性化的系统和方法。 通常，识别器可用于对输入数据进行分类。 在个性化期间，向识别器提供特定于用户，实体或格式的样本，以提高特定用户，实体或格式的性能。 在个性化过程中可以利用偏置正则化来维持非用户特定输入的识别器性能。 在一个方面，正则化可以偏向识别器的原始参数，使得识别器在个性化期间不被过度修改。