公开(公告)号：US11778223B2

公开(公告)日：2023-10-03

申请号：US17406845

申请日：2021-08-19

申请人： Wentao Liu ,  Yuanhao Yu ,  Yang Wang ,  Juwei Lu ,  Xiaolin Wu ,  Jin Tang

发明人： Wentao Liu ,  Yuanhao Yu ,  Yang Wang ,  Juwei Lu ,  Xiaolin Wu ,  Jin Tang

IPC分类号： H04N19/59 ,  H04N19/51 ,  H04N19/184 ,  H04N19/136

CPC分类号： H04N19/51 ,  H04N19/136 ,  H04N19/184

摘要： A method, device and computer-readable medium for generating a super-resolution version of a compressed video stream. By leveraging the motion information and residual information in compressed video streams, described examples are able to skip the time-consuming motion-estimation step for most frames and make the most use of the SR results of key frames. A key frame SR module generates SR versions of I-frames and other key frames of a compressed video stream using techniques similar to existing multi-frame approaches to VSR. A non-key frame SR module generates SR version of the non-key inter frames between these key frames by making use of motion information and residual information used to encode the inter frames in the compressed video stream.

公开(公告)号：US11430138B2

公开(公告)日：2022-08-30

申请号：US17102114

申请日：2020-11-23

申请人： Zhixiang Chi ,  Rasoul Mohammadi Nasiri ,  Zheng Liu ,  Jin Tang ,  Juwei Lu

发明人： Zhixiang Chi ,  Rasoul Mohammadi Nasiri ,  Zheng Liu ,  Jin Tang ,  Juwei Lu

IPC分类号： G06T7/246 ,  G06N3/08

摘要： Systems and methods for multi-frame video frame interpolation. Higher-order motion modeling, such as cubic motion modeling, achieves predictions of intermediate optical flow between multiple interpolated frames, assisted by relaxation of the constraints imposed by the loss function used in initial optical flow estimation. A temporal pyramidal optical flow refinement module performs coarse-to-fine refinement of the optical flow maps used to generate the intermediate frames, focusing a proportionally greater amount of refinement attention to the optical flow maps for the high-error middle frames. A temporal pyramidal pixel refinement module performs coarse-to-fine refinement of the generated intermediate frames, focusing a proportionally greater amount of refinement attention to the high-error middle frames. A generative adversarial network (GAN) module calculates a loss function for training the neural networks used in the optical flow estimation module, temporal pyramidal optical flow refinement module, and/or temporal pyramidal pixel refinement module.

公开(公告)号：US20140056511A1

公开(公告)日：2014-02-27

申请号：US13592148

申请日：2012-08-22

申请人： Juwei Lu ,  Bradley Scott Denney

发明人： Juwei Lu ,  Bradley Scott Denney

IPC分类号： G06K9/62

CPC分类号： G06K9/6256 ,  G06K9/4676

摘要： Systems and methods for generating a visual vocabulary build a plurality of visual words via unsupervised learning on set of features of a given type; decompose one or more visual words to a collection of lower-dimensional buckets; generate labeled image representations based on the collection of lower dimensional buckets and labeled images, wherein labels associated with an image are associated with a respective representation of the image; and iteratively select a sub-collection of buckets from the collection of lower-dimensional buckets based on the labeled image representations, wherein bucket selection during any iteration after an initial iteration is based at least in part on feedback from previously selected buckets.

摘要翻译： 用于产生视觉词汇的系统和方法通过给定类型的特征集合上的无监督学习来构建多个视觉词; 将一个或多个视觉词分解为低维度桶的集合; 基于较低维度桶和标记图像的集合生成标记图像表示，其中与图像相关联的标签与图像的相应表示相关联; 并且基于标记的图像表示从所述低维度桶的集合中迭代地选择桶的子集合，其中在初始迭代之后的任何迭代期间的桶选择至少部分地基于来自先前选择的桶的反馈。

公开(公告)号：US20090222388A1

公开(公告)日：2009-09-03

申请号：US12313193

申请日：2008-11-17

申请人： Wei Hua ,  Xiangrong Chen ,  Ryan Crabb ,  Jonathan Cook ,  Juwei Lu

发明人： Wei Hua ,  Xiangrong Chen ,  Ryan Crabb ,  Jonathan Cook ,  Juwei Lu

IPC分类号： G06N3/08 ,  G06K9/00 ,  G06N5/02 ,  G06F15/18 ,  G06F17/00

CPC分类号： G06N5/02 ,  G06K9/00778

摘要： The present invention is directed to a computer automated method of selectively identifying a user specified behavior of a crowd. The method comprises receiving video data but can also include audio data and sensor data. The video data contains images a crowd. The video data is processed to extract hierarchical human and crowd features. The detected crowd features are processed to detect a selectable crowd behavior. The selected crowd behavior detected is specified by a configurable behavior rule. Human detection is provided by a hybrid human detector algorithm which can include Adaboost or convolutional neural network. Crowd features are detected using textual analysis techniques. The configurable crowd behavior for detection can be defined by crowd behavioral language.

摘要翻译： 本发明涉及一种选择性地识别用户指定的人群行为的计算机自动化方法。 该方法包括接收视频数据，但也可以包括音频数据和传感器数据。 视频数据包含人群中的图像。 视频数据被处理以提取分层的人和人群特征。 检测到的人群特征被处理以检测可选择的人群行为。 检测到的所选群体行为由可配置行为规则指定。 人类检测由可以包括Adaboost或卷积神经网络的混合人体检测器算法提供。 使用文本分析技术检测人群特征。 用于检测的可配置人群行为可以由人群行为语言来定义。

公开(公告)号：US20080144892A1

公开(公告)日：2008-06-19

申请号：US11948026

申请日：2007-11-30

申请人： Juwei Lu ,  Mohanaraj Thiyagarajah ,  Hui Zhou

发明人： Juwei Lu ,  Mohanaraj Thiyagarajah ,  Hui Zhou

IPC分类号： G06K9/00 ,  G06F15/00

CPC分类号： G06K9/00228

摘要： Converting a digital image from color to gray-scale. In one example embodiment, a method for converting a digital image from color to gray-scale is disclosed. First, an unconverted pixel having red, green, and blue color channels is selected from the color digital image. Next, the red color channel of the pixel is multiplied by α. Then, the green color channel of the pixel is multiplied by β. Next, the blue color channel of the pixel is multiplied by γ. Then, the results of the three multiplication operations are added together to arrive at a gray-scale value for the pixel. Finally, these acts are repeated for each remaining unconverted pixel of the color digital image to arrive at a gray-scale digital image. In this example method, α+β+≈1 and α>β.

摘要翻译： 将数字图像从颜色转换为灰度。 在一个示例实施例中，公开了一种将数字图像从彩色转换成灰度级的方法。 首先，从彩色数字图像中选择具有红色，绿色和蓝色通道的未转换像素。 接下来，将像素的红色通道乘以α。 然后，像素的绿色通道乘以β。 接下来，将像素的蓝色通道乘以伽马。 然后，将三个乘法运算的结果相加在一起，得到像素的灰度值。 最后，对于彩色数字图像的每个剩余的未转换像素重复这些动作，以得到灰度数字图像。 在此示例中，alpha + beta +≈1和alpha>β。

公开(公告)号：US08977041B2

公开(公告)日：2015-03-10

申请号：US13592148

申请日：2012-08-22

申请人： Juwei Lu ,  Bradley Scott Denney

发明人： Juwei Lu ,  Bradley Scott Denney

IPC分类号： G06K9/62

CPC分类号： G06K9/6256 ,  G06K9/4676

摘要： Systems and methods for generating a visual vocabulary build a plurality of visual words via unsupervised learning on set of features of a given type; decompose one or more visual words to a collection of lower-dimensional buckets; generate labeled image representations based on the collection of lower dimensional buckets and labeled images, wherein labels associated with an image are associated with a respective representation of the image; and iteratively select a sub-collection of buckets from the collection of lower-dimensional buckets based on the labeled image representations, wherein bucket selection during any iteration after an initial iteration is based at least in part on feedback from previously selected buckets.

摘要翻译： 用于产生视觉词汇的系统和方法通过给定类型的特征集合上的无监督学习来构建多个视觉词; 将一个或多个视觉词分解为低维度桶的集合; 基于较低维度桶和标记图像的集合生成标记图像表示，其中与图像相关联的标签与图像的相应表示相关联; 并且基于标记的图像表示从所述低维度桶的集合中迭代地选择桶的子集合，其中在初始迭代之后的任何迭代期间的桶选择至少部分地基于来自先前选择的桶的反馈。

公开(公告)号：US20140015855A1

公开(公告)日：2014-01-16

申请号：US13550357

申请日：2012-07-16

申请人： Bradley Scott Denney ,  Juwei Lu ,  Dariusz T. Dusberger ,  Sholeh Forouzan

发明人： Bradley Scott Denney ,  Juwei Lu ,  Dariusz T. Dusberger ,  Sholeh Forouzan

IPC分类号： G09G5/00

CPC分类号： G06K9/6218 ,  G06F16/258 ,  G06F16/355 ,  G06F16/367 ,  G06F17/2785 ,  G06T11/001 ,  G06T11/60 ,  G09G5/14

摘要： Systems and methods for clustering descriptors in a space of visual descriptors to generate augmented visual descriptors in an augmented space that includes semantic information, wherein the augmented space of the augmented descriptors includes both visual descriptor-to-descriptor dissimilarities and semantic label-to-label dissimilarities; and cluster the augmented visual descriptors in the augmented space based at least in part on a dissimilarity measure between augmented visual descriptors in the augmented descriptor space.

摘要翻译： 用于在视觉描述符的空间中聚类描述符以在扩展空间中生成包括语义信息的增强视觉描述符的系统和方法，其中所述扩展描述符的扩充空间包括视觉描述符到描述符的不相似性和语义标签到标号 不相似 并且至少部分地基于增强的描述符空间中的增强的视觉描述符之间的不相似度量度来在增强空间中聚集增强的视觉描述符。

公开(公告)号：US20130279881A1

公开(公告)日：2013-10-24

申请号：US13451436

申请日：2012-04-19

申请人： Juwei Lu ,  Bradley Denney

发明人： Juwei Lu ,  Bradley Denney

IPC分类号： H04N9/80

CPC分类号： G06T7/0081 ,  G06F17/30849 ,  G06K9/00751 ,  G06K9/00765 ,  G11B27/031 ,  G11B27/105 ,  H04N5/91 ,  H04N21/2668 ,  H04N21/8549

摘要： Systems and methods for summarizing a video assign frames in a video to at least one of two or more groups based on a topic, generate a respective first similitude measurement for the frames in a group relative to the other frames in the group based on a feature, rank the frames in a group relative to one or more other frames in the group based on the respective first similitude measurement of the respective frames, and select a frame from each group as a most-representative frame based on the respective rank of the frames in a group relative to the other frames in the group.

摘要翻译： 用于总结视频的视频的系统和方法基于主题将视频中的帧分配给两个或更多个组中的至少一个，基于特征生成组中相对于组中的其他帧的帧的相应的第一相似度测量 基于相应帧的相应的第一相似度测量对组中的一个或多个其他帧进行排序，并且基于帧的相应等级从每个组中选择一个帧作为最具代表性的帧 在组中相对于组中的其他帧。

公开(公告)号：US07983480B2

公开(公告)日：2011-07-19

申请号：US11750099

申请日：2007-05-17

申请人： Juwei Lu ,  Hui Zhou ,  Mohanaraj Thiyagarajah

发明人： Juwei Lu ,  Hui Zhou ,  Mohanaraj Thiyagarajah

IPC分类号： G06K9/36

CPC分类号： G06K9/00234

摘要： A method and system for scanning a digital image for detecting the representation of an object, such as a face, and for reducing memory requirements of the computer system performing the image scan. One example method includes identifying an original image and downsamples the original image in an x-dimension and in a y-dimension to obtain a downsampled image that requires less storage space than the original digital image. A first scan is performed of the downsampled image to detect the representation of an object within the downsampled image. Then, the original digital image is divided into at least two image blocks, where each image block contains a portion of the original digital image. A second scan is then performed of each of the image blocks to detect the representation of the object within the image blocks.

摘要翻译： 一种用于扫描数字图像以检测诸如面部的对象的表示以及用于减少执行图像扫描的计算机系统的存储器需求的方法和系统。 一个示例性方法包括识别原始图像并且以x维度和y维度对原始图像进行下采样以获得比原始数字图像更少的存储空间的下采样图像。 执行下采样图像的第一扫描以检测下采样图像中的对象的表示。 然后，原始数字图像被分成至少两个图像块，其中每个图像块包含原始数字图像的一部分。 然后对每个图像块执行第二扫描以检测图像块内的对象的表示。

公开(公告)号：US20080100716A1

公开(公告)日：2008-05-01

申请号：US11838750

申请日：2007-08-14

申请人： Guoyi Fu ,  Juwei Lu

发明人： Guoyi Fu ,  Juwei Lu

IPC分类号： H04N5/228

CPC分类号： H04N5/23248

摘要： Methods for estimating a point spread function of a blurred digital image. One example method includes capturing gyro data during an image exposure time, deriving gyro samples from the gyro data at predetermined gyro sampling times, calculating a motion vector field of the image at each gyro sampling time, approximating an overall image scene motion path by averaging motion paths of selected pixels in the image, and estimating the point spread function from the approximated overall image scene motion path.

摘要翻译： 用于估计模糊数字图像的点扩散函数的方法。 一个示例性方法包括在图像曝光时间期间捕获陀螺仪数据，在预定陀螺仪采样时间从陀螺仪数据导出陀螺仪样本，在每个陀螺仪采样时间计算图像的运动矢量场，通过平均运动近似整个图像场景运动路径 图像中的所选像素的路径，以及从近似的整体图像场景运动路径估计点扩散函数。