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

WO2022133832A1 APPARATUS AND METHOD FOR PREDICTIVE COMPUTER MODELLING 审中-公开
公开(公告)号：WO2022133832A1
公开(公告)日：2022-06-30
申请号：PCT/CN2020/138724
申请日：2020-12-23
申请人： HUAWEI TECHNOLOGIES CO., LTD.
发明人： LI, Minne , YANG, Mengyue , LIU, Furui , CHEN, Xu , CHEN, Zhitang , WANG, Jun
IPC分类号： G06F16/901
摘要： An apparatus for generating a predictive model of a real-world environment in which a plurality of objects change state. The apparatus comprises one or more processors configured to receive state information representing one or more changes in state of a plurality of training objects over respective periods of time and train a neural network to estimate for each training object a variant factor (u). The variant factor is estimated in dependence on a series of states of the training object over a period of time as defined by the state information. The variant factor (u) is estimated by repeatedly optimising for a loss function between (i) a first state of the training object at a given time as defined by the state information and (ii) a second state of the training object as estimated for the given time by a first model in dependence on a state of the training object as defined by the state information for a time prior to the given time.

2. 发明申请

WO2020073928A1 INTER PREDICTION METHOD AND APPARATUS 审中-公开
公开(公告)号：WO2020073928A1
公开(公告)日：2020-04-16
申请号：PCT/CN2019/110194
申请日：2019-10-09
申请人： HUAWEI TECHNOLOGIES CO., LTD.
发明人： CHEN, Xu , ZHENG, Jianhua
IPC分类号： H04N19/176 , H04N19/50
摘要： Inter prediction method and apparatus. The method includes: obtaining initial motion information of at least two sub-blocks of a current picture block (S601); determining motion information of the at least two sub-blocks based on the initial motion information of the at least two sub-blocks and positions of the at least two sub-blocks (S602); determining motion information of the current picture block based on the motion information of the at least two sub-blocks (S603); and determining a prediction block of the current picture block based on the motion information of the current picture block (S604).

3. 发明申请

WO2019061537A1 METHOD AND APPARATUS FOR DERIVATION OF TEMPORAL MOTION DATA CANDIDATES IN VIDEO CODING 审中-公开
公开(公告)号：WO2019061537A1
公开(公告)日：2019-04-04
申请号：PCT/CN2017/105133
申请日：2017-09-30
申请人： HUAWEI TECHNOLOGIES CO., LTD.
发明人： CHEN, Xu , AN, Jicheng , ZHENG, Jianhua
IPC分类号： H04N19/94
CPC分类号： H04N19/52 , H04N19/503 , H04N19/94
摘要： A method for deriving of one or more temporal motion data (TMD) candidates for a current block in video coding or video decoding is disclosed. It is determined whether a first prediction block at a primary TMD position is available, wherein the primary TMD position is a bottom right neighboring position outside a co-located block；and motion data of the primary TMD position and/or motion data of a secondary TMD position may be selected as the TMD candidates when the first prediction block at the primary TMD position is available and when the number of current available motion data candidates for an inter-prediction candidate list is smaller than a target number, wherein the secondary TMD position is a central position inside the co-located block. This method allows for providing more meaningful motion data candidates for inclusion in the inter-prediction candidate list for the current block, thus the coding performance of video data can be improved.

4. 发明申请

WO2020073973A1 INTER PREDICTION METHOD AND APPARATUS 审中-公开
公开(公告)号：WO2020073973A1
公开(公告)日：2020-04-16
申请号：PCT/CN2019/110488
申请日：2019-10-10
申请人： HUAWEI TECHNOLOGIES CO., LTD.
发明人： CHEN, Xu , ZHENG, Jianhua
IPC分类号： H04N19/52
摘要： Methods and apparatuses for inter prediction. The method comprises: constructing a merge candidate list of a current picture block; adding HMVP candidate motion information to the merge candidate list, in a case that a quantity of candidate motion information in the merge candidate list is less than a maximum candidate quantity of the merge candidate list minus N, wherein N is a positive integer; obtaining a merge index of the current picture block; obtaining motion information of the current picture block based on the merge candidate list and the merge index; and determining a predicted block of the current picture block based on the motion information of the current picture block.

5. 发明申请

WO2020069652A1 CANDIDATE MV CONSTRUCTION METHOD FOR HMVP MODE 审中-公开
公开(公告)号：WO2020069652A1
公开(公告)日：2020-04-09
申请号：PCT/CN2019/106260
申请日：2019-09-17
申请人： HUAWEI TECHNOLOGIES CO., LTD.
发明人： CHEN, Xu , ZHENG, Jianhua
IPC分类号： H04N19/52
摘要： A method and apparatus for generating a history-based motion vector prediction (HMVP) candidate list configured to decode a video image includes obtaining a first motion vector (MV) of a to-be-decoded image block in the HMVP candidate list, parsing offset identifier information in a received bit stream, and obtaining a second MV based on the first MV and the offset identifier information of the first MV. The offset identifier information includes offset information of the first MV configured to provide a horizontal offset value and a vertical offset value that modify the first MV for obtaining the second MV. The second MV is then added to the HMVP candidate list.

6. 发明公开

EP3783888A1 VIDEO ENCODING AND DECODING METHOD AND APPARATUS 审中-公开
公开(公告)号：EP3783888A1
公开(公告)日：2021-02-24
申请号：EP18919173.7
申请日：2018-05-16
申请人： HUAWEI TECHNOLOGIES CO., LTD.
发明人： CHEN, Xu , AN, Jicheng , ZHENG, Jianhua
IPC分类号： H04N19/105
摘要： Embodiments of this application disclose a method for obtaining a motion vector, including: determining a reference block of a to-be-processed block, where the reference block and the to-be-processed block have a preset temporal or spatial correlation, the reference block has an initial motion vector and one or more preset motion vector offsets, the initial motion vector of the reference block is obtained based on a predicted motion vector of the reference block, and a prediction block of the reference block is obtained based on the initial motion vector and the one or more preset motion vector offsets; and using the initial motion vector of the reference block as a predicted motion vector of the to-be-processed block.

7. 发明授权

EP2469787B1 Method and device for preventing network attacks 失效
标题翻译：用于防止网络攻击的方法和装置
公开(公告)号：EP2469787B1
公开(公告)日：2014-06-18
申请号：EP09848395.1
申请日：2009-10-16
申请人： Huawei Technologies Co., Ltd.
发明人： SONG, Xuan , CHEN, Xu , JIANG, Sheng
IPC分类号： H04L29/06 , H04L29/12
CPC分类号： H04L61/103 , H04L63/1458 , H04L63/1466 , H04L69/16 , H04L69/22

8. 发明公开

EP2136504A1 TRANSMITTING AND RECEIVING METHOD, APPARATUS AND SYSTEM FOR SECURITY POLICY OF MULTICAST SESSION 失效
标题翻译：发送 - EMPFANGSVERFAHREN，VORRICHTUNG系统FUER DIE SICHERHEITSPOLITIK EINER MULTICAST-SITZUNG
公开(公告)号：EP2136504A1
公开(公告)日：2009-12-23
申请号：EP08783908.0
申请日：2008-08-07
申请人： Huawei Technologies Co., Ltd.
发明人： CHEN, Xu
IPC分类号： H04L9/08 , H04L29/06
CPC分类号： H04L9/083 , H04L9/0891 , H04L63/065 , H04L63/20 , H04L65/4076 , H04L65/608
摘要： A method, apparatus, and system for sending and receiving the security policy of multicast sessions are provided. The method for sending the security policy of multicast sessions includes: after a Datagram Transport Layer Security (DTLS) session is set up between a sender and a receiver, receiving a security policy request sent from the receiver, constructing a security policy response according to the security policy, multiplexing the security policy response and Secure Real-Time Transport Protocol (SRTP) multicast session data, and sending the multiplexed data to the receiver. An in-band mode is adopted to complete security policy negotiation, which solves the issue that the key negotiation solution in the prior art cannot realize safe negotiation about the security policy of SRTP multicast sessions.
摘要翻译：提供了一种用于发送和接收多播会话的安全策略的方法，装置和系统。发送组播会话的安全策略的方法包括：在发送方和接收方之间建立数据报传输层安全（DTLS）会话后，接收从接收者发送的安全策略请求，根据安全策略，多路复用安全策略响应和安全实时传输协议（SRTP）多播会话数据，并将复用的数据发送到接收器。采用带内方式完成安全策略协商，解决了现有技术关键协商解决方案无法实现SRTP组播会话安全策略安全协商的问题。

9. 发明公开

EP2063566A1 AN ENCRYPTED KEY UPDATING SYSTEM, METHOD THEREOF AND A TRANSMITTING TERMINAL AND A RECEIVING TERMINAL 审中-公开
标题翻译： VERSCHLÜSSELTESSCHLÜSSELAKTUALISIERUNGSSYSTEM，VERFAHRENDAFÜRSOWIESENDEENDGERÄTEMEMANGSENDGERÄT
公开(公告)号：EP2063566A1
公开(公告)日：2009-05-27
申请号：EP08715107.2
申请日：2008-02-28
申请人： Huawei Technologies Co., Ltd.
发明人： CHEN, Xu
IPC分类号： H04L9/18
CPC分类号： H04L9/065 , H04L9/0891 , H04L9/12 , H04L2209/601
摘要： A key update system includes an encrypting module and a multiplexing module at a transmitting side, and a de-multiplexing module and a session key generating module at a receiving side. The encrypting module is configured to encrypt a multimedia stream with a session key to form a cipher text for transmission to the multiplexing module; the multiplexing module is configured to multiplex a key update packet and the cipher text to form a composite data stream, and send the composite data stream to the de-multiplexing module, the key update packet including parameters needed for generating the session key; the de-multiplexing module is configured to separate the key update packet and the cipher text in a received composite data stream; and the session key generating module is configured to decrypt the key update packet to recover the parameters needed for generating the session key, from which to generate an update session key. The disclosure also provides a transmitting side and a receiving side for implementing key update, a key update method, methods for transmitting and receiving a key update packet.
摘要翻译：密钥更新系统包括发送侧的加密模块和复用模块，以及接收侧的解复用模块和会话密钥生成模块。所述加密模块被配置为利用会话密钥对多媒体流进行加密以形成用于传输到复用模块的密文; 复用模块被配置为复用密钥更新分组和密文以形成复合数据流，并将复合数据流发送到解复用模块，密钥更新分组包括生成会话密钥所需的参数; 解复用模块被配置为在所接收的复合数据流中分离密钥更新分组和密文; 并且会话密钥生成模块被配置为解密密钥更新分组以恢复生成会话密钥所需的参数，从其生成更新会话密钥。本公开还提供了用于实现密钥更新的发送侧和接收侧，密钥更新方法，用于发送和接收密钥更新分组的方法。

10. 发明公开

EP4210327A1 INTRA FRAME PREDICTION METHOD AND DEVICE 审中-实审
公开(公告)号：EP4210327A1
公开(公告)日：2023-07-12
申请号：EP21871636.3
申请日：2021-09-26
申请人： Huawei Technologies Co., Ltd.
发明人： YANG, Haitao , SONG, Nan , CHEN, Xu , MA, Xiang , CHEN, Huanbang , ZHAO, Yin
IPC分类号： H04N19/11
摘要： This application provides an intra prediction method and an apparatus. This application relates to the field of artificial intelligence (AI)-based video or picture compression technologies, and in particular, to the field of neural network-based video compression technologies. The method includes: obtaining respective intra prediction modes or texture distributions of P reconstructed picture blocks in a surrounding region of a current block; obtaining, based on the respective intra prediction modes or texture distributions of the P reconstructed picture blocks, Q priori candidate intra prediction modes of the current block and Q probability values; obtaining, based on M probability values corresponding to M priori candidate intra prediction modes, M weighting factors corresponding to the M priori candidate intra prediction modes; separately performing intra prediction based on the M priori candidate intra prediction modes to obtain M predicted values; and obtaining a predicted value of the current block based on a weighted summation of the M predicted values and the corresponding M weighting factors. This application can improve accuracy of intra prediction, reduce an error of intra prediction, and improve RDO efficiency of intra prediction.

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