-
31.发明公开
公开(公告)号：EP4394597A1
公开(公告)日：2024-07-03
申请号：EP22925691.2
申请日：2022-11-17
发明人： YANG, Kun , WANG, Cilang , TANG, Xi , YANG, Guangdong
CPC分类号： G06F2209/50920130101 , G06F9/505 , G06F9/5072
摘要： Embodiments of the present application provide a graphics processing unit resource management method and apparatus, a device, a computer readable storage medium, and a computer program product. The method provided by the embodiments of the present application relates to the field of cloud technology and cloud games. For a plurality of application processes simultaneously running on a same graphics processing unit, a resource allocation scheme is determined in real time in consideration of remaining available resources of the application processes in historical resource allocation on the basis of the number of currently available resources of the application processes among graphics processing unit resources, thereby implementing efficient allocation of graphics processing unit resources. By means of the method of the embodiments of the present application, graphics processing unit resources can be reasonably allocated according to resource requirements of application processes, thereby avoiding the effect of contention between a plurality of application processes, and improving the utilization rate of graphics processor resources.
-
32.发明公开
公开(公告)号：EP4393657A1
公开(公告)日：2024-07-03
申请号：EP23844982.1
申请日：2023-05-11
发明人： WANG, Shuai , ZHENG, Yu , ZHOU, Qinqin , CHI, Wanchao
CPC分类号： B62D57/032 , B25J9/16
摘要： Embodiments of this application provide a method, an apparatus, and a device for controlling a legged robot, a legged robot, a computer-readable storage medium, and a computer program product. The legged robot includes a base and at least two robotic legs. Each of the robotic legs includes at least one joint. The method includes: determining a first expected moving trajectory and a second expected moving trajectory corresponding to the legged robot in response to determining that each robotic leg of the legged robot falls to contact a plane, the first expected moving trajectory indicating an expected moving trajectory of a center of mass of the legged robot moving in a direction of gravity and an expected moving trajectory of a change in a tilt angle of the legged robot, and the second expected moving trajectory indicating an expected moving trajectory of a foot end of each robotic leg; and controlling, based on a dynamic model corresponding to the legged robot, the first expected moving trajectory, and the second expected moving trajectory, an action of each joint after the legged robot contacts the plane.
-
公开(公告)号：EP4391560A1
公开(公告)日：2024-06-26
申请号：EP22925684.7
申请日：2022-11-15
发明人： ZHAO, Yuanyuan , REN, Jingjing , LI, Chen
IPC分类号： H04N21/44
CPC分类号： H04N19/132 , H04N19/167 , H04N19/17 , H04N19/124 , G06V20/40 , G06V10/26 , G06V10/7715
摘要： The present application relates to the technical field of video processing. Disclosed are a video processing method and apparatus, a device and a storage medium. The method comprises: inputting a video to be processed and a video mask into a coding model for feature extraction to obtain a first video feature vector and a second video feature vector (202);determining an index distribution of a first quantized feature vector in a discrete latent space formed on the basis of the first quantized feature vector (204); determining a second quantized feature vector in the discrete latent space on the basis of the second video feature vector and the index distribution(206);and inputting the first quantized feature vector and the second quantized feature vector into a decoding model to obtain a reconstructed video (208).
-
34.发明公开
公开(公告)号：EP4390660A1
公开(公告)日：2024-06-26
申请号：EP22913504.1
申请日：2022-09-13
发明人： REN, Zimu , LI, Dongsheng
IPC分类号： G06F7/483
摘要： A multi-input floating point number processing method executed by a computer device, the method comprising: obtaining a plurality of floating point numbers to be processed corresponding to a target task, and respectively extracting an exponent value of an exponential part and a mantissa value of a mantissa part in each floating point number (S102); sorting the plurality of floating point numbers according to the exponent value of each floating point number to obtain a sorting result, and on the basis of the sorting result, respectively allocating to each floating point number a shifter from a plurality of shifters having different preset digits (S104); for each floating point number, performing shift processing on the mantissa value of the corresponding floating point number by means of the shifter allocated to the floating point number to obtain a shift result (S106);and on the basis of each shift result, determining a floating point number processing result corresponding to the target task (S 108).
-
公开(公告)号：EP4390461A1
公开(公告)日：2024-06-26
申请号：EP22928296.7
申请日：2022-11-17
发明人： GAO, Xingwei
摘要： A data processing method and apparatus, and a device (1000) and a storage medium, which are mainly applied to the field of electronic maps. The method comprises: acquiring first actual observation data (22, 32, 42), second actual observation data (24), the actual position of a target reference station (11, 23) and the actual positions (33, 43) of N satellites (S101, S201); determining an initially estimated position (26, 35, 45) of a terminal (21, 31, 41, 71) according to the first actual observation data (22, 32, 42) and the actual positions (33, 43) of the N satellites (S102, S202); according to the second actual observation data (24), the actual position of the target reference station (11, 23) and the actual positions (33, 43) of the N satellites, determining a first system positioning error (27) corresponding to the initially estimated position (26, 35, 45) of the terminal (21, 31, 41, 71) (S103, S203); and adjusting the initially estimated position (26, 35, 45) of the terminal (21, 31, 41, 71) according to the first system positioning error and the first actual observation data (22, 32, 42), so as to obtain a target position (28) of the terminal (21, 31, 41, 71) (S104). Therefore, the position positioning accuracy of the terminal (21, 31, 41, 71) can be improved.
-
公开(公告)号：EP4387239A1
公开(公告)日：2024-06-19
申请号：EP22906077.7
申请日：2022-10-26
发明人： LI, Zhicheng
IPC分类号： H04N21/234 , H04N19/114 , H04N19/177
CPC分类号： H04N21/442 , H04N21/234 , H04N21/24 , H04N19/177 , H04N21/44 , H04N19/114 , H04N19/164 , H04N19/50 , H04N19/136 , H04N19/172 , H04N19/105 , H04N19/107
摘要： The present application provides a video encoding method and apparatus, an electronic device, and a storage medium. The method comprises: encoding a first image frame in a video clip to be encoded into a video key frame; and for a plurality of image frames after the first image frame, respectively performing the following operations in sequence: if it is determined that a network condition of a receiving end meets a transceiving condition, encoding, on the basis of a frame type corresponding to a current image frame to be encoded in a first encoding structure, the current image frame to be encoded; and if it is determined that the network condition does not meet the transceiving condition, encoding, on the basis of a frame type corresponding to the current image frame to be encoded in a second encoding structure, the current image frame to be encoded; wherein the number of bidirectionally predictive-coded frames in the second encoding structure is less than the number of the bidirectionally predictive-coded frames in the first encoding structure.
-
公开(公告)号：EP4386748A1
公开(公告)日：2024-06-19
申请号：EP23861664.3
申请日：2023-09-15
发明人： SU, Guinan , WU, Gao , LI, Zhifeng , LIU, Wei , CUI, Yushi
IPC分类号： G10L21/10 , G06T13/40 , G10L25/03 , G06T15/00 , H04N21/439
CPC分类号： Y02D10/00 , G06T13/40 , G06T15/00 , G06V40/16 , G10L21/10 , G10L25/03 , H04N21/439 , H04N21/44
摘要： The embodiments of this application disclose a video generation method and apparatus, and a computer-readable storage medium. The method includes: obtaining initial status data of a virtual object and audio data; extracting an audio feature from the audio data, and encoding the audio data to obtain an encoded feature; fusing the audio feature and the encoded feature to obtain a fused audio feature of the virtual object; obtaining updated status data of the virtual object according to the fused audio feature and the initial status data; and generating video data according to the updated status data and the audio data. Therefore, accuracy of matching a target part in video data with audio data is improved, accuracy of video generation for the virtual object is improved, and efficiency of video generation for the virtual object is further improved.
-
38.发明授权
公开(公告)号：EP3933829B1
公开(公告)日：2024-06-12
申请号：EP20860732.5
申请日：2020-07-13
IPC分类号： G10L21/0208 , G10L15/16 , G10L15/02 , G10L15/06
CPC分类号： G10L21/0208 , G10L15/02 , G10L15/063 , G10L15/16
-
公开(公告)号：EP3674968B1
公开(公告)日：2024-06-05
申请号：EP18874465.0
申请日：2018-10-17
IPC分类号： G06N3/048 , G06V10/20 , G06V10/32 , G06V10/82 , G06V40/18 , G06N3/045 , G06V10/764 , G06V10/774 , G06N3/084
CPC分类号： G06N3/084 , G06V40/197 , G06V10/20 , G06V10/32 , G06V2201/0320220101 , G06V10/82 , G06V10/764 , G06V10/774 , G06N3/048 , G06N3/045 , Y02T10/40
-
公开(公告)号：EP4365774A1
公开(公告)日：2024-05-08
申请号：EP22836674.6
申请日：2022-06-13
CPC分类号： G06T3/4053 , G06T7/33 , G06T2207/1001620130101 , G06T2207/1005620130101 , G06T2207/3002420130101 , G06T2207/2008420130101
摘要： The present disclosure discloses a microscope-based super-resolution method and apparatus, device and medium, and relates to the field of image processing. The method includes: acquiring an input image and at least one auxiliary image, an overlap area existing between a display area of the auxiliary image and a target area in the input image, and the input image and the auxiliary image being both images of a first resolution degree (302); registering the input image and the auxiliary image to obtain a registered image (304); acquiring a high-resolution feature based on the registered image, the high-resolution feature configured to represent image feature of a second resolution degree in the target area (306); and reconstructing the high-resolution feature to obtain a target image of the second resolution degree (308). The above-mentioned method is used for processing a microscope image sequence to acquire and reconstruct a high-resolution image good in effect.
-
-
-
-
-
-
-
-
-
搜索结果
1195 条记录 (耗时 0.034 秒)
