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

WO2019033380A1 SLIMMING OF NEURAL NETWORKS IN MACHINE LEARNING ENVIRONMENTS 审中-公开
公开(公告)号：WO2019033380A1
公开(公告)日：2019-02-21
申请号：PCT/CN2017/097977
申请日：2017-08-18
申请人： INTEL CORPORATION , YAN, Shoumeng , LI, Jianguo , LIU, Zhuang
发明人： YAN, Shoumeng , LI, Jianguo , LIU, Zhuang
IPC分类号： G06N3/08
摘要： A mechanism is described for facilitating slimming of neural networks in machine learning environments. A method of embodiments, as described herein, includes learning a first neural network associated with machine learning processes to be performed by a processor of a computing device, where learning includes analyzing a plurality of channels associated with one or more layers of the first neural network. The method may further include computing a plurality of scaling factors to be associated with the plurality of channels such that each channel is assigned a scaling factor, wherein each scaling factor to indicate relevance of a corresponding channel within the first neural network. The method may further include pruning the first neural network into a second neural network by removing one or more channels of the plurality of channels having low relevance as indicated by one or more scaling factors of the plurality of scaling factors assigned to the one or more channels.

2. 发明申请

WO2020061884A1 COMPOSITE BINARY DECOMPOSITION NETWORK 审中-公开
公开(公告)号：WO2020061884A1
公开(公告)日：2020-04-02
申请号：PCT/CN2018/107886
申请日：2018-09-27
申请人： INTEL CORPORATION , LI, Jianguo , CHEN, Yurong
发明人： LI, Jianguo , CHEN, Yurong , WANG, Zheng
IPC分类号： G06N3/06
摘要： Embodiments are directed to a composite binary decomposition network. An embodiment of a computer-readable storage medium includes executable computer program instructions for transforming a pre-trained first neural network into a binary neural network by processing layers of the first neural network in a composite binary decomposition process, where the first neural network having floating point values representing weights of various layers of the first neural network. The composite binary decomposition process includes a composite operation to expand real matrices or tensors into a plurality of binary matrices or tensors, and a decompose operation to decompose one or more binary matrices or tensors of the plurality of binary matrices or tensors into multiple lower rank binary matrices or tensors.

3. 发明申请

WO2014139105A1 IMPROVED TECHNIQUES FOR THREE-DIMENSIONAL IMAGE EDITING 审中-公开
标题翻译：改进的三维图像编辑技术
公开(公告)号：WO2014139105A1
公开(公告)日：2014-09-18
申请号：PCT/CN2013/072544
申请日：2013-03-13
申请人： INTEL CORPORATION , DING, Dayong , DU, Yangzhou , LI, Jianguo
发明人： DING, Dayong , DU, Yangzhou , LI, Jianguo
IPC分类号： G06F11/00
CPC分类号： G06T11/60 , G06T15/00 , H04N13/128
摘要： Techniques for three-dimensional (3D) image editing are described. In one embodiment, for example, an apparatus may comprise a processor circuit and a 3D graphics management module, and the 3D graphics management module maybe operable by the processor circuit to determine modification information for a first sub-image in a 3D image comprising the first sub -image and a second sub -image, modify the first sub- image based on the modification information for the first sub-image, determine modification information for the second sub -image based on the modification information for the first sub-image, and modify the second sub-image based on the modification information for the second sub-image. Other embodiments are described and claimed.
摘要翻译：描述三维（3D）图像编辑技术。在一个实施例中，例如，设备可以包括处理器电路和3D图形管理模块，并且3D图形管理模块可以由处理器电路操作以确定3D图像中的第一子图像的修改信息，所述3D图像包括第一子图像和第二子图像，基于第一子图像的修改信息修改第一子图像，基于第一子图像的修改信息确定第二子图像的修改信息，以及基于第二子图像的修改信息修改第二子图像。描述和要求保护其他实施例。

4. 发明申请

WO2013063765A1 OBJECT DETECTION USING EXTENDED SURF FEATURES 审中-公开
标题翻译：使用扩展的特征的对象检测
公开(公告)号：WO2013063765A1
公开(公告)日：2013-05-10
申请号：PCT/CN2011/081642
申请日：2011-11-01
申请人： INTEL CORPORATION , LI, Jianguo , ZHANG, Yimin
发明人： LI, Jianguo , ZHANG, Yimin
IPC分类号： G06K9/00
CPC分类号： G06K9/6217 , G06K9/4671 , G06K9/6257
摘要： Systems, apparatus and methods are described including generating gradient images from an input image, where the gradient images include gradient images created using 2D filter kernels. Feature descriptors are then generated from the gradient images and object detection performed by applying the descriptors to a boosting cascade classifier that includes logistic regression base classifiers.
摘要翻译：描述了系统，装置和方法，其包括从输入图像生成渐变图像，其中梯度图像包括使用2D滤波器内核创建的渐变图像。然后，通过将描述符应用于包括逻辑回归基类分类器的升压级联分类器，从梯度图像和对象检测中生成特征描述符。

5. 发明申请

WO2012139275A1 OBJECT OF INTEREST BASED IMAGE PROCESSING 审中-公开
标题翻译：基于兴趣的图像处理对象
公开(公告)号：WO2012139275A1
公开(公告)日：2012-10-18
申请号：PCT/CN2011/072601
申请日：2011-04-11
申请人： INTEL CORPORATION , LI, Qiang , LI, Wenlong , WANG, Peng , WANG, Tao , DU, Yangzhou , LI, Jianguo
发明人： LI, Qiang , LI, Wenlong , WANG, Peng , WANG, Tao , DU, Yangzhou , LI, Jianguo
IPC分类号： G06T9/00
CPC分类号： H04N7/15 , H04N19/115 , H04N19/167 , H04N19/17 , H04N19/59
摘要： An apparatus, a method and a system are provided, wherein the system includes an encoding engine to encode and/or compress one or more objects of interest within individual image frames with higher bit densities than the bit density employed to encode and/or compress their background. The system may further include a context engine to identify a region of interest including at least a part of the one or more objects of interest, and scale the region of interest within individual image frames to emphasize the objects of interest.
摘要翻译：提供了一种装置，方法和系统，其中所述系统包括编码引擎，以在具有比用于编码和/或压缩它们的比特密度更高的比特密度的单个图像帧内编码和/或压缩一个或多个感兴趣对象背景。系统还可以包括上下文引擎，以识别包括所述一个或多个感兴趣对象的至少一部分的感兴趣区域，以及在各个图像帧内对感兴趣的区域进行缩放以强调感兴趣的对象。

6. 发明申请

WO2010148539A1 TECHNIQUES TO DETECT VIDEO COPIES 审中-公开
标题翻译：检测视频复制的技术
公开(公告)号：WO2010148539A1
公开(公告)日：2010-12-29
申请号：PCT/CN2009/000716
申请日：2009-06-26
申请人： INTEL CORPORATION , WANG, Tao , LI, Jianguo , LI, Wenlong , ZHANG, Yimin
发明人： WANG, Tao , LI, Jianguo , LI, Wenlong , ZHANG, Yimin
IPC分类号： G06T1/00 , G06F17/30
CPC分类号： G06F17/30814 , G06K9/00335 , G06K9/00758
摘要： Some embodiments include a video copy detection approach based on speeded up robust features (SURF) trajectory building, local sensitive hash (LSH) indexing, and spatial-temporal-scale registration. First, interesting points' trajectories are extracted by SURF. Next, an efficient voting based spatial-temporal-scale registration approach is applied to estimate the optimal transformation parameters (shift and scale) and achieve the final video copy detection results by propagations of video segments in both spatial-temporal and scale directions. To speed up the detection speed, local sensitive hash (LSH) indexing is used to index trajectories for fast queries of candidate trajectories.
摘要翻译：一些实施例包括基于加速鲁棒特征（SURF）轨迹建立，局部敏感散列（LSH）索引和空间 - 时间尺度注册的视频拷贝检测方法。首先，SURF提取了有趣的点的轨迹。接下来，应用有效的基于投票的空间 - 时间尺度注册方法来估计最佳变换参数（移位和缩放），并通过视频片段在空间 - 时间和尺度方向上的传播来实现最终的视频拷贝检测结果。为了加快检测速度，本地敏感散列（LSH）索引用于索引候选轨迹快速查询的轨迹。

7. 发明申请

WO2019090506A1 SCENE TEXT DETECTOR FOR UNCONSTRAINED ENVIRONMENTS 审中-公开
公开(公告)号：WO2019090506A1
公开(公告)日：2019-05-16
申请号：PCT/CN2017/109885
申请日：2017-11-08
申请人： INTEL CORPORATION , CHENG, Wenhua , YAO, Anbang , WANG, Libin , CAI, Dongqi , LI, Jianguo , CHEN, Yurong
发明人： CHENG, Wenhua , YAO, Anbang , WANG, Libin , CAI, Dongqi , LI, Jianguo , CHEN, Yurong
IPC分类号： G06K9/20 , G06K9/72
CPC分类号： G06K9/325 , G06K9/6267 , G06K9/72
摘要： A semiconductor package apparatus (20), comprising: one or more substrates (21); and logic (22) coupled to the one or more substrates (21), the logic (22) coupled to the one or more substrates (21) to: apply a trained scene text detection network to an image to identify a core text region, a supportive text region (31), and a background region of the image, and detect text in the image based on the identified core text region and supportive text region (32).

8. 发明申请

WO2017096570A1 VISUAL RECOGNITION USING DEEP LEARNING ATTRIBUTES 审中-公开
标题翻译：用深度学习属性进行视觉识别
公开(公告)号：WO2017096570A1
公开(公告)日：2017-06-15
申请号：PCT/CN2015/096882
申请日：2015-12-10
申请人： INTEL CORPORATION , LI, Jianguo , LUO, Jianwei , CHEN, Yurong
发明人： LI, Jianguo , LUO, Jianwei , CHEN, Yurong
IPC分类号： G06K9/66
CPC分类号： G06K9/4628 , G06K9/6269 , G06K9/66 , G06N3/04 , G06N3/08
摘要： A processing device for performing visual recognition using deep learning attributes and method for performing the same are described. In one embodiment, a processing device comprises: an interface to receive an input image； and a recognition unit coupled to the interface and operable to perform visual object recognition on the input image, where the recognition unit has an extractor to extract region proposals from the input image, a convolutional neural network (CNN) to compute features for each extracted region proposal, the CNN being operable to create a soft-max layer output, a cross region pooling unit operable to perform pooling of the soft-max layer output to create a set of attributes of the input image, and an image classifier operable to perform image classification based on the attributes of the input image.
摘要翻译：描述了使用深度学习属性执行视觉识别的处理设备和用于执行该处理的方法。在一个实施例中，处理设备包括：接口，用于接收输入图像;以及识别单元，其耦合到接口并且可操作以在输入图像上执行视觉对象识别，其中识别单元具有提取器以从输入图像，用于计算每个提取的区域提议的特征的卷积神经网络（CNN），所述CNN可操作以创建软 - 最大层输出;交叉区域合并单元，可操作以执行合并所述软 - 最大层输出以创建输入图像的一组属性，以及图像分类器，可操作用于基于输入图像的属性执行图像分类。

9. 发明申请

WO2013097160A1 INTERACTIVE MEDIA SYSTEMS 审中-公开
标题翻译：互动媒体系统
公开(公告)号：WO2013097160A1
公开(公告)日：2013-07-04
申请号：PCT/CN2011/084970
申请日：2011-12-30
申请人： INTEL CORPORATION , WANG, Tao , SONG, Qing Jian Edwin , LI, Jianguo , DU, Yangzhou , LI, Wenlong , ZHANG, Yimin
发明人： WANG, Tao , SONG, Qing Jian Edwin , LI, Jianguo , DU, Yangzhou , LI, Wenlong , ZHANG, Yimin
IPC分类号： H04N21/4415
CPC分类号： H04N21/4415 , G06T11/00 , H04N21/4223 , H04N21/4316 , H04N21/44008 , H04N21/4532 , H04N21/4788 , H04N2101/00 , H04N2209/041
摘要： Generally this disclosure describes interactive media methods and systems. A method may include capturing an image, detecting at least one face in the image, determining an identity and expression corresponding to the at least one face, generating an icon for the at least one face based on the corresponding expression, and displaying the icon on a video monitor.
摘要翻译：通常本公开描述了交互式媒体方法和系统。一种方法可以包括拍摄图像，检测图像中的至少一个脸部，确定与至少一个脸部相对应的身份和表情，基于相应的表情生成至少一张脸部的图标，并将图标显示在视频监视器。

10. 发明申请

WO2013086734A1 REDUCED IMAGE QUALITY FOR VIDEO DATA BACKGROUND REGIONS 审中-公开
标题翻译：视频数据的降低图像质量背景区域
公开(公告)号：WO2013086734A1
公开(公告)日：2013-06-20
申请号：PCT/CN2011/084118
申请日：2011-12-16
申请人： INTEL CORPORATION , WANG, Peng , ZHANG, Yimin , LI, Qiang, Eric , LI, Jianguo , XU, Lin
发明人： WANG, Peng , ZHANG, Yimin , LI, Qiang, Eric , LI, Jianguo , XU, Lin
IPC分类号： G06K9/34 , G06K9/36 , G06K9/60
CPC分类号： G06K9/00268 , H04N19/117 , H04N19/167 , H04N19/17 , H04N19/85
摘要： Systems, apparatus, articles, and methods are described including operations to detect a face based at least in part on video data. A region of interest and a background region may be determined based at least in part on the detected face. The background region may be modified to have a reduced image quality.
摘要翻译：描述了系统，装置，物品和方法，其包括至少部分地基于视频数据检测脸部的操作。可以至少部分地基于检测到的面部来确定感兴趣的区域和背景区域。背景区域可以被修改为具有降低的图像质量。

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