-
公开(公告)号：US09230159B1
公开(公告)日：2016-01-05
申请号：US14100595
申请日：2013-12-09
Applicant: Google Inc.
CPC classification number: G06K9/00342 , G06K9/3241
Abstract: This disclosure generally relates to systems and methods that facilitate employing exemplar Histogram of Oriented Gradients Linear Discriminant Analysis (HOG-LDA) models along with Localizer Hidden Markov Models (HMM) to train a classification model to classify actions in videos by learning poses and transitions between the poses associated with the actions in a view of a continuous state represented by bounding boxes corresponding to where the action is located in frames of the video.
Abstract translation: 本公开通常涉及有助于使用定向梯度直方图线性判别分析（HOG-LDA）模型与定位器隐马尔可夫模型（HMM）一起使用的系统和方法来训练分类模型以通过学习姿态和视频之间的转换来分类视频中的动作 与由行动位于视频帧中的对应框所表示的连续状态的视图相关联的姿势。
-
公开(公告)号：US09177208B2
公开(公告)日：2015-11-03
申请号：US13633062
申请日：2012-10-01
Applicant: Google Inc.
Inventor： Rahul Sukthankar , Jay Yagnik
CPC classification number: G06K9/00744 , G06F17/3079 , G06F17/3082 , G06K9/00718 , G06T9/00
Abstract: A volume identification system identifies a set of unlabeled spatio-temporal volumes within each of a set of videos, each volume representing a distinct object or action. The volume identification system further determines, for each of the videos, a set of volume-level features characterizing the volume as a whole. In one embodiment, the features are based on a codebook and describe the temporal and spatial relationships of different codebook entries of the volume. The volume identification system uses the volume-level features, in conjunction with existing labels assigned to the videos as a whole, to label with high confidence some subset of the identified volumes, e.g., by employing consistency learning or training and application of weak volume classifiers. The labeled volumes may be used for a number of applications, such as training strong volume classifiers, improving video search (including locating individual volumes), and creating composite videos based on identified volumes.
Abstract translation: 体积识别系统识别一组视频中的每一个中的一组未标记的时空体积，每个体积表示不同的对象或动作。 音量识别系统进一步为每个视频确定表征整个音量的一组音量级特征。 在一个实施例中，特征基于码本并且描述卷的不同码本条目的时间和空间关系。 音量识别系统使用音量级特征，结合分配给整个视频的现有标签，以高度置信的方式标识所识别的体积的一些子集，例如通过采用一致性学习或训练和应用弱音量分类器 。 标记的卷可以用于许多应用，例如训练强大的分类器，改进视频搜索（包括定位各个卷），以及基于识别的卷创建复合视频。
-
公开(公告)号：US09619521B1
公开(公告)日：2017-04-11
申请号：US14142976
申请日：2013-12-30
Applicant: Google Inc.
Inventor： Rahul Sukthankar , Jay Yagnik
CPC classification number: G06N99/005 , G06F17/3079
Abstract: A segmentation annotation technique for media items is disclosed herein. Given a weakly labeled media item, spatiotemporal masks may be generated for each of the concepts with which it is labeled. Segments may be ranked by the likelihood that they correspond to a given concept. The ranked concept segments may be utilized to train a classifier that, in turn, may be used to classify untagged or new media items.
-
公开(公告)号：US09087242B2
公开(公告)日：2015-07-21
申请号：US13633067
申请日：2012-10-01
Applicant: Google Inc.
Inventor： Rahul Sukthankar , Jay Yagnik
IPC: G06F3/0481 , G11B27/34 , G06K9/00
CPC classification number: G06K9/00744 , G06F17/3079 , G06F17/3082 , G06K9/00718 , G06T9/00
Abstract: A volume identification system identifies a set of unlabeled spatio-temporal volumes within each of a set of videos, each volume representing a distinct object or action. The volume identification system further determines, for each of the videos, a set of volume-level features characterizing the volume as a whole. In one embodiment, the features are based on a codebook and describe the temporal and spatial relationships of different codebook entries of the volume. The volume identification system uses the volume-level features, in conjunction with existing labels assigned to the videos as a whole, to label with high confidence some subset of the identified volumes, e.g., by employing consistency learning or training and application of weak volume classifiers.The labeled volumes may be used for a number of applications, such as training strong volume classifiers, improving video search (including locating individual volumes), and creating composite videos based on identified volumes.
Abstract translation: 体积识别系统识别一组视频中的每一个中的一组未标记的时空体积，每个体积表示不同的对象或动作。 音量识别系统进一步为每个视频确定表征整个音量的一组音量级特征。 在一个实施例中，特征基于码本并且描述卷的不同码本条目的时间和空间关系。 音量识别系统使用音量级特征，结合分配给整个视频的现有标签，以高度置信的方式标识所识别的体积的一些子集，例如通过采用一致性学习或训练和应用弱音量分类器 。 标记的卷可以用于许多应用，例如训练强大的分类器，改进视频搜索（包括定位各个卷），以及基于识别的卷创建复合视频。
-
-
-
Search Results
4
records
(took 0.011 seconds)
