公开(公告)号：US20250078369A1

公开(公告)日：2025-03-06

申请号：US18953942

申请日：2024-11-20

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

Inventor： Guohao LI ,  Jiachen LIU ,  Xinyan XIAO

IPC: G06T11/60 ,  G06F40/174 ,  G06F40/186 ,  G06F40/40 ,  G06T5/60 ,  G06T5/70

Abstract: A method is provided that includes: obtaining an editing instruction input by a user in a current round of a dialogue and history dialogue information in at least one history round of the dialogue, wherein the history dialogue information comprises a history dialogue text and at least one history image; determining a source image to be edited from the at least one history image based on the editing instruction and the history dialogue information; and editing the source image to generate a target image based on the editing instruction.

公开(公告)号：US20240338862A1

公开(公告)日：2024-10-10

申请号：US18749461

申请日：2024-06-20

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

Inventor： Jiachen LIU ,  Xinyan XIAO ,  Hua WU ,  Guohao LI ,  Wei LI ,  Hong ZHU ,  Qiaoqiao SHE ,  Yajuan LV

IPC: G06T11/00 ,  G06F40/35

CPC classification number: G06T11/00 ,  G06F40/35

Abstract: A method is provided that includes: obtaining current dialogue data; determining a requirement type of the user in the current round of dialogue based on the current dialogue data; in response to the requirement type being an image processing requirement, determining an action sequence for implementing the image processing requirement; executing the action sequence to generate a target image; and generating response data corresponding to the user input data based on the target image.

公开(公告)号：US20240362493A1

公开(公告)日：2024-10-31

申请号：US18770122

申请日：2024-07-11

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

Inventor： Yixuan SHI ,  Wei LI ,  Jiachen LIU ,  Xinyan XIAO

IPC: G06N3/092 ,  G06N3/045 ,  G06N3/0475 ,  G06N3/09 ,  G06T11/00

CPC classification number: G06N3/092 ,  G06N3/045 ,  G06N3/0475 ,  G06N3/09 ,  G06T11/00

Abstract: A method is provided that includes: obtaining a first Text-to-Image model and a pre-trained reward model, wherein the first Text-to-Image model is used to generate a corresponding image based on input text, and the pre-trained reward model is used to score a data pair composed of the input text and the corresponding generated image; and adjusting the parameters of the first Text-to-Image model based on the pre-trained reward model and a reinforcement learning policy to obtain a second Text-to-Image model.

公开(公告)号：US20230004589A1

公开(公告)日：2023-01-05

申请号：US17577561

申请日：2022-01-18

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

Inventor： Wenhao WU ,  Wei LI ,  Xinyan XIAO ,  Jiachen LIU

IPC: G06F16/34 ,  G06F40/51 ,  G06F40/56

Abstract: The present disclosure provides a summary generation model training method and apparatus, a device and a storage medium, and relates to the field of computer technologies, and in particular, to the field of artificial intelligence such as natural language processing and deep learning. The summary generation model training method includes: acquiring a document representation corresponding to a document sample; constructing, based on the document representation, a summary representation corresponding to the document representation, the summary representation including a positive summary representation and a negative summary representation; and constructing a total contrastive loss function based on the document representation, the positive summary representation and the negative summary representation, and training a summary generation model based on the total contrastive loss function. The present disclosure may improve accuracy of the summary generation model.

公开(公告)号：US20220138435A1

公开(公告)日：2022-05-05

申请号：US17572930

申请日：2022-01-11

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

Inventor： Zhe HU ,  Zhiwei CAO ,  Jiachen LIU

IPC: G06F40/40

Abstract: The disclosure provides a method for generating a text. The method includes: obtaining a coding sequence of a first text by coding the first text; obtaining a controllable attribute of a second text to be generated; predicting a hidden state of the second text based on the coding sequence of the first text and the controllable attribute of the second text; and obtaining a second text corresponding to the first text by decoding the coding sequence of the first text based on the hidden state of the second text.