公开(公告)号：US11496744B2

公开(公告)日：2022-11-08

申请号：US17541092

申请日：2021-12-02

Applicant: Beijing Bytedance Network Technology Co., Ltd. ,  Bytedance Inc.

Inventor： Li Zhang ,  Kai Zhang ,  Hongbin Liu ,  Yang Wang ,  Yue Wang

IPC: H04N19/52 ,  H04N19/137 ,  H04N19/105 ,  H04N19/132 ,  H04N19/176

Abstract: A method of video processing includes determining, for a conversion between a current block of a video and a bitstream representation of the video, an operation associated with a list of motion candidates based on a condition related to a characteristic of the current block. The list of motion candidates is constructed for a coding technique or based on information from previously processed blocks of the video. The method also includes performing the conversion based on the determining.

公开(公告)号：US20210026860A1

公开(公告)日：2021-01-28

申请号：US16980897

申请日：2018-09-07

Applicant: BEIJING BYTEDANCE NETWORK TECHNOLOGY CO., LTD.

Inventor： Yang Wang ,  Ziniu Hu ,  Qu Peng ,  Hang Li

IPC: G06F16/2457 ,  G06F16/248 ,  G06N5/00 ,  G06F17/18

Abstract: The embodiment of the present application discloses a method and a device for generating a ranking model. A specific embodiment of the method includes: acquiring a sample set, executing following training steps: for the samples in the sample set, inputting the query information, the first position document and the second position document in the sample into an initial model, respectively obtaining scores of the input documents, and determining a target value of the sample based on the obtained scores, a clicked bias of a first position and an unclicked bias of a second position, updating the initial model based on the target value of each sample; determining whether the initial model is completely trained; and in response to determining that the initial model is completely trained, determining the updated initial model as the ranking model.

公开(公告)号：US11776535B2

公开(公告)日：2023-10-03

申请号：US17885965

申请日：2022-08-11

Applicant: BEIJING BYTEDANCE NETWORK TECHNOLOGY CO., LTD.

Inventor： He Zhang ,  Hang Li ,  Yang Wang

IPC: G10L15/18 ,  G10L15/183

CPC classification number: G10L15/1815 ,  G10L15/183

Abstract: A semantic understanding method and apparatus, and a device and a storage medium are provided. The method includes: acquiring a recognition character string that matches speech information; acquiring, from an entity vocabulary library, at least one entity vocabulary respectively corresponding to each recognition character in the recognition character string; and according to a situation of each entity vocabulary hitting the recognition character string, determining a matching entity vocabulary as a semantic understanding result of the speech information. By means of the method, insofar as a completely matching entity vocabulary is not acquired, a matching entity vocabulary can still be determined according to an entity vocabulary library, and semantic information of speech is thus accurately understood; and the method also has relatively high fault tolerance for situations such as wrong words, added words, and omitted words, such that the semantic understanding accuracy of speech information is improved.