公开(公告)号：US10929683B2

公开(公告)日：2021-02-23

申请号：US16362580

申请日：2019-03-22

Applicant: BEIJING BAIDU NETCOM SCIENCE TECHNOLOGY CO., LTD.

Inventor： Sijiao Chen ,  Shuai Jiang ,  Shiliang Diao ,  Haijin Liang ,  Yu Luo

IPC: G06K9/00 ,  G06K9/46 ,  G10L17/06 ,  G10L17/00

Abstract: A video processing method, apparatus and device. The method includes: acquiring a video to be processed; performing audio processing on a plurality of frames of audios in the video to obtain a first processing result corresponding to each frame of audio; performing image processing on a plurality of frames of images in the video to obtain a second processing result corresponding to each frame of image; determining an entity in the video according to the first processing result and the second processing result; acquiring association information of the entity according to index information of the entity; generating first text information used for describing each frame of image, according to the second processing result corresponding to each frame of image and association information of the entity; and generating a first type of target article according to the plurality of frames of images and first text information of each frame of image.

公开(公告)号：US20220083874A1

公开(公告)日：2022-03-17

申请号：US17532968

申请日：2021-11-22

Applicant: BEIJING BAIDU NETCOM SCIENCE TECHNOLOGY CO., LTD.

Inventor： Donghai Bian ,  Weihua Peng ,  Yu Luo ,  Shuai Jiang

IPC: G06N5/02 ,  G06K9/62 ,  G06F40/279 ,  G06F16/957

Abstract: A method and apparatus for training a search model, and a method and apparatus for searching for a target object are provided. The method may comprise: obtaining, based on sample query information, a first sample data set used to train the search model. The method may further comprise: determining, based on a pre-constructed knowledge base, additional query information associated with the sample query information. In addition, the method may further comprise: obtaining, based on the additional query information, a second sample data set used to train the search model, wherein the second sample data set is different from the first sample data set. Moreover, the method may further comprise: training the search model based on the sample query information, the first sample data set, and the second sample data set.