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. 发明公开

EP3410362A1 METHOD AND APPARATUS FOR NEXT TOKEN PREDICTION BASED ON PREVIOUSLY OBSERVED TOKENS 有权
公开(公告)号：EP3410362A1
公开(公告)日：2018-12-05
申请号：EP18175232.0
申请日：2018-05-30
申请人： HERE Global B.V.
发明人： Pietrobon, Davide , Lewis, Andrew , Macfarlane, Jane , Berry, Robert
IPC分类号： G06Q10/04 , G06Q50/30
CPC分类号： G08G1/0129 , G06F17/30241 , G06K9/6218 , G06Q10/04 , G06Q50/30 , G08G1/0133
摘要： An approach is provided for next token prediction based on previously observed tokens. The approach involves receiving an observed time series of tokens, wherein each of the tokens represents an observed data pattern. The approach also involves adding a most recent token from the observed time series of tokens into a variable token set. The approach further involves processing a historical token set to determine a historical token sequence comprising the variable token set followed by a next token. The approach further involves recursively adding a next most recent token from the observed time series of tokens into the variable token set for processing until the next token following the variable token set in the determined historical token sequence is unique or meets a target number of possible predictions. The approach further involves presenting the next token as a predicted next token of the observed time series of tokens.

2. 发明公开

EP3379427A1 ORDER CLUSTERING METHOD AND DEVICE, AND MALICIOUS INFORMATION REJECTING METHOD AND DEVICE 审中-公开
公开(公告)号：EP3379427A1
公开(公告)日：2018-09-26
申请号：EP16865698.1
申请日：2016-11-09
申请人： Alibaba Group Holding Limited
发明人： LI, Jiuxi
IPC分类号： G06F17/30
CPC分类号： G06Q30/0635 , G06F17/30 , G06F17/30705 , G06F17/30887 , G06F17/30997 , G06K9/6218 , G06Q30/0185 , H04L63/1483
摘要： The present application relates to the field of network communications technologies, and in particular, to order clustering and malicious information combating method and apparatus. The order clustering method includes: generating a unique order ID and order content based on information about an order; converting the order content to an order vector; and performing a clustering operation based on the unique order ID and the order vector to obtain an order clustering result. After the clustering, it can be further determined, according to content of an order in the cluster, whether the cluster is a malicious information cluster. If the cluster is a malicious information cluster, orders in the entire cluster are correspondingly processed. In the method and the apparatus in the present application, an order in an electronic transaction can be automatically classified, a new cluster can be obtained through analysis based on real-time content of the order, and orders in an entire cluster can be uniformly processed. It improves efficiency of combating malicious information.

3. 发明授权

EP2984624B1 SUPERPIXEL GENERATION WITH IMPROVED SPATIAL COHERENCY 有权
公开(公告)号：EP2984624B1
公开(公告)日：2018-08-22
申请号：EP14711955.6
申请日：2014-03-18
申请人： Thomson Licensing
发明人： JACHALSKY, Joern , RESO, Matthias , ROSENHAHN, Bodo
IPC分类号： G06T7/10
CPC分类号： G06K9/6218 , G06K9/6227 , G06K9/6272 , G06T7/11 , G06T7/12 , G06T7/187 , G06T2207/20016 , G06T2207/20116
摘要： A method for generating superpixels for an image comprising pixels and an apparatus configured to perform the method. A clustering stage clusters the pixels into initial clusters. A determination stage then selects spatially coherent parts of the clusters and determines unconnected fragments of the clusters. Pixels of these unconnected fragments are iteratively assigned to adjacent clusters by an assignment stage using a contour evolution process.

4. 发明公开

EP3361414A1 TOPOGRAPHIC DATA MACHINE LEARNING METHOD AND SYSTEM 审中-公开
公开(公告)号：EP3361414A1
公开(公告)日：2018-08-15
申请号：EP18154908.0
申请日：2018-02-02
申请人： Ordnance Survey Limited
发明人： Sargent, Isabel , Hare, Jonathon
IPC分类号： G06K9/00 , G06K9/62
CPC分类号： G06K9/0063 , G06K9/6218 , G06K9/6256 , G06K9/6273 , G06K9/6282 , G06K9/726 , G06N3/0454 , G06N3/08
摘要： Embodiments of the invention apply machine learning techniques for image recognition and classification to the processing of topographic imagery, in order to permit more accurate and detailed topographic representations of an area to be obtained. In particular, in one embodiment a machine learning system is trained with existing topographic imagery and corresponding topographic data relating to a particular area, so that the machine learning system is then able to relate actual physical topographical features to their topographic representations in existing data. Having been so trained, the machine learning system may then be used to process topographic imagery data of the same area to determine new topographic details thereof for incorporation into the topographic data.

5. 发明授权

EP3184051B1 CLUSTERING, NOISE REDUCTION AND VISUALIZATION METHOD FOR DOPPLER ULTRASOUND IMAGES 有权
公开(公告)号：EP3184051B1
公开(公告)日：2018-08-15
申请号：EP16205350.8
申请日：2016-12-20
申请人： AmCad BioMed Corporation
发明人： CHEN, Argon , CHEN, Jia-Jiun , SHEN, Yu-Han , LIU, Cheng-Hsien
IPC分类号： A61B8/08 , A61B8/06 , A61B8/00 , G01S15/89 , G06F17/30
CPC分类号： G01S7/52077 , A61B8/06 , A61B8/08 , A61B8/0891 , A61B8/463 , A61B8/468 , A61B8/469 , A61B8/488 , A61B8/5207 , A61B8/5223 , G01S7/52036 , G01S7/52053 , G01S7/52071 , G01S7/52074 , G01S15/8988 , G06F17/3025 , G06K9/42 , G06K9/4642 , G06K9/4652 , G06K9/4661 , G06K9/6215 , G06K9/6218 , G06K9/623 , G06K9/628
摘要： The invention provides an analysis system for Doppler ultrasound image includes: a capture device, a processing device, and an output device. The capture device obtains a plurality of Doppler ultrasound images. The processing device arranges the color value in each pixel of the images based on the time domain, and obtains a reference sequence through a referencing method, furthermore performs a clustering method to obtain a plurality of correlation coefficient values, then uses a clustering and noise reducing method to classify into a primary pulsatile signal, a secondary pulsatile signal, and a noise signal, finally annotates the primary pulsatile signal, a secondary pulsatile signal, and a noise signal with different color values, respectively. The output device displays a plurality of visualized pulsatile ultrasound images for visualization.

6. 发明公开

EP3357002A1 SEMI-AUTOMATIC LABELLING OF DATASETS 审中-公开
公开(公告)号：EP3357002A1
公开(公告)日：2018-08-08
申请号：EP16795403.1
申请日：2016-10-03
申请人： Tractable Ltd.
发明人： DALYAC, Alexandre , RANCA, Razvan , HOGAN, Robert , MCALEESE-PARK, Nathaniel John , CHATFIELD, Ken
IPC分类号： G06K9/62
CPC分类号： G06K9/3241 , G06K9/6218 , G06K9/6259 , G06K9/6263 , G06K2209/23 , G06N5/046 , G06N7/08 , G06N99/005
摘要： An unlabelled or partially labelled target dataset is modelled with a machine learning model for classification (or regression). The target dataset is processed by the machine learning model; a subgroup of the target dataset is prepared for presentation to a user for labelling or label verification; label verification or user re-labelling or user labelling of the subgroup is received; and the updated target dataset is re-processed by the machine learning model. User labelling or label verification combined with modelling an unclassified or partially classified target dataset with a machine learning model aims to provide efficient labelling of an unlabelled component of the target dataset.

7. 发明公开

EP3349165A1 RECOGNITION METHOD AND APPARATUS FOR USER RELATIONSHIP, STORAGE MEDIUM, AND SERVER 审中-公开
公开(公告)号：EP3349165A1
公开(公告)日：2018-07-18
申请号：EP15905253.9
申请日：2015-12-25
申请人： Baidu Online Network Technology (Beijing) Co., Ltd
发明人： WANG, Chao , YU, Hao , GUO, Yang
IPC分类号： G06Q30/02
CPC分类号： H04L67/22 , G06F17/30946 , G06K9/6218 , G06K9/6269 , G06K9/6276 , G06K9/6282 , G06Q30/02 , G06Q30/0201 , G06Q50/01 , H04W4/02 , H04W8/22 , H04W8/26
摘要： Embodiments of the present disclosure disclose a method and an apparatus for recognizing a user relationship, a storage medium and a server. The method comprises: acquiring a wireless network address and coordinate information of each of networked user terminals; training a position-setting classifier based on the wireless network address and the coordinate information of each of known users having a known set position coordinate ; determining a set position wireless network address for each of unknown users having an unknown set position coordinate based on the position-setting classifier; and marking a relationship of at least two users based on a common relationship of the set position wireless network address. In the present disclosure, based on networked information of user terminals, set position wireless network addresses of unknown users are determined by training classifiers to mark relationships among users. In this way, relationships among the users, such as families, friends or colleagues, may be precisely distinguished, the problem that acquaintances collaboratively cheat to acquire illegal economic incomes in commercial promotion activities of the mobile Internet thereby causing the loss of promotion activities cost is solved, and the effect of Internet promotion activities is enhanced.

8. 发明授权

EP3050290B1 METHOD AND APPARATUS FOR VIDEO ANTI-SHAKING 有权
公开(公告)号：EP3050290B1
公开(公告)日：2018-06-27
申请号：EP14848246.6
申请日：2014-09-18
申请人： Nokia Technologies Oy
发明人： JIANG, An , CUI, Feng , ZHU, Haitao , WANG, Yan
IPC分类号： G06T7/20 , G06T7/238 , G06T5/00 , G06T5/10 , G06T5/50 , H04N5/232 , G06K9/62 , G06T7/223
CPC分类号： G06T5/50 , G06K9/6218 , G06T5/002 , G06T5/10 , G06T7/223 , G06T7/238 , G06T2207/10016 , G06T2207/20021 , G06T2207/20064 , G06T2207/20201 , H04N5/23254 , H04N5/23277
摘要： Embodiments of the present invention relate to method and apparatus for video anti-shaking. There is disclosed a method for use in a video anti-shaking, especially for real-time digital image stabilization, the method comprises: dividing at least one part of a current frame of a video into a plurality of regions (S301); performing a progressive local motion detection for each of the plurality of regions to determine a local motion vector for the region (S302, S303, S304); and determining a global motion vector for the current frame based on the local motion vectors for the plurality of regions (S305). There is also described a corresponding apparatus and user equipment. With embodiments of the present invention, the video anti-shaking processing can be done without any additional hardware. Moreover, embodiments of the present invention may quickly and effectively determine a motion vector and corresponding compensation.

9. 发明公开

EP3198858A4 METHOD AND SYSTEM OF LENS SHADING COLOR CORRECTION USING BLOCK MATCHING 有权
公开(公告)号：EP3198858A4
公开(公告)日：2018-05-09
申请号：EP15843307
申请日：2015-08-20
申请人： INTEL CORP
发明人： PALIY DMYTRO
IPC分类号： H04N9/07 , H04N5/217 , H04N9/04 , H04N17/02
CPC分类号： H04N9/646 , B41J2/473 , G06K9/4642 , G06K9/4652 , G06K9/6218 , H04N5/145 , H04N5/217 , H04N5/3572 , H04N9/045
摘要： A system, article, and method to provide lens shading color correction using block matching.

10. 发明公开

EP3306493A1 CLASSIFICATION METHOD, DEVICE AND SYSTEM FOR PROGRAM FILE 有权
标题翻译：程序文件的分类方法，装置和系统
公开(公告)号：EP3306493A1
公开(公告)日：2018-04-11
申请号：EP16887739.7
申请日：2016-12-07
申请人： Huawei Technologies Co., Ltd.
发明人： LIU, Zhenhua
IPC分类号： G06F17/30
CPC分类号： G06F21/53 , G06F17/30 , G06F21/55 , G06K9/6218
摘要： A program file classification method, a program file classification apparatus, and a program file classification system are disclosed. The system uses an agent program in a client and a sandbox server to obtain behavior information corresponding to at least two behaviors executed by a program file at runtime. Each piece of behavior information includes a behavior identifier and a path related during execution of a corresponding behavior. A classification server performs normalization processing on the path in each piece of behavior information, where the normalization processing is used to reduce path diversity; generates a feature vector according to at least two pieces of behavior information obtained after the path normalization processing; and determines, according to the feature vector, a category to which the program file belongs. According to the classification method, the classification apparatus, and the classification system in embodiments of the present invention, normalization processing is performed on the path, so that randomness of a path obtained after the normalization processing is reduced, so as to improve a program file classification effect and accordingly reduce a workload in identifying a malicious program file.
摘要翻译：公开了一种程序文件分类方法，程序文件分类装置和程序文件分类系统。该系统使用客户端和沙箱服务器中的代理程序来获取与在运行时由程序文件执行的至少两个行为相对应的行为信息。每条行为信息包括在对应行为的执行期间相关的行为标识符和路径。分类服务器对每条行为信息中的路径进行归一化处理，归一化处理用于减少路径分集; 根据路径归一化处理后的至少两条行为信息生成特征向量; 并根据该特征向量确定该程序文件所属的类别。根据本发明实施例中的分类方法，分类装置和分类系统，对路径进行归一化处理，以减少归一化处理后得到的路径的随机性，从而改善程序文件分类效果，并相应减少识别恶意程序文件的工作量。

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