公开(公告)号：US20240127259A1

公开(公告)日：2024-04-18

申请号：US18395240

申请日：2023-12-22

Applicant: MaShang Consumer Finance Co., Ltd.

Inventor： Fuming GENG ,  Haiying WU ,  Ning JIANG

IPC: G06Q30/015

CPC classification number: G06Q30/015

Abstract: Embodiments of the present application provide a user question labeling method and apparatus. By obtaining dialogue data generated by a manual customer service system, determining a target quick response statement corresponding to a target raised question, and then determining, in combination with a correspondence between a preset quick response statement and a user intention label, a target intention label corresponding to the target quick response statement, so as to obtain a correspondence between the target raised question and the target intention label, labeling user intention for a user question is thus achieved and then a generated user intention recognition sample set is applied to a user intention recognition model in an intelligent customer service system.

公开(公告)号：US20240241879A1

公开(公告)日：2024-07-18

申请号：US18561872

申请日：2022-11-14

Applicant: MASHANG CONSUMER FINANCE CO., LTD.

Inventor： Dengpeng ZUO ,  Haiying WU ,  Linchengxi ZENG

IPC: G06F16/23

CPC classification number: G06F16/2365 ,  G06F16/2358

Abstract: A data processing method. In the method, a target transaction change log of each database sub-library among a plurality of database sub-libraries is acquired and parsed. The target transaction change log is a most recently generated transaction change log in the database sub-library. If an event type indicated by the target transaction change log is a Data Definition Languages (DDL) event, information of the DDL event, which has been recorded in the target transaction change log, is recorded in a target event slot corresponding to the DDL event.

公开(公告)号：US20250022315A1

公开(公告)日：2025-01-16

申请号：US18896742

申请日：2024-09-25

Applicant: MaShang Consumer Finance Co., Ltd.

Inventor： Xingfu ZHAO ,  Ning JIANG ,  Haiying WU

IPC: G06V40/20 ,  G06V10/75 ,  G06V40/16

Abstract: The present application relates to the field of image processing technology, specifically to a lip movement detection method and apparatus, a computer-readable storage medium and an electronic device, solving the problem of weak generalization ability and poor robustness of traditional lip movement detection method. The lip movement detection method provided in an embodiment of the present application determines a lip movement detection result of a user based on a reference interlabial distance of the user in a first image frame, where the reference interlabial distance is determined based on a correspondence between an interlabial distance and a reference value for interlabial distance, thus the reference interlabial distance is a relative value. That is, the reference interlabial distance is not easily affected by factors such as shooting angle and shooting distance, thereby improving the robustness of the lip movement detection method.

公开(公告)号：US20240135956A1

公开(公告)日：2024-04-25

申请号：US18395253

申请日：2023-12-22

Applicant: MaShang Consumer Finance Co., Ltd.

Inventor： Chun WANG ,  Dingheng ZENG ,  Haiying WU ,  Xunyi ZHOU ,  Ning JIANG

IPC: G10L25/57 ,  G06V40/16 ,  G10L15/24

CPC classification number: G10L25/57 ,  G06V40/165 ,  G06V40/168 ,  G10L15/24

Abstract: The application provides a method and an apparatus for measuring speech-image synchronicity, and a method and an apparatus for training a model, where the method for measuring speech-image synchronicity includes: acquiring a speech segment and an image segment of a video, where there is a correspondence between the speech segment and the image segment in the video; processing the speech segment and the image segment to obtain a speech feature of the speech segment and a visual feature of the image segment; and determining, according to the speech feature of the speech segment and the visual feature of the speech segment, whether there is synchronicity between the speech segment and the image segment, where the synchronicity is used for characterizing matching between a sound in the speech segment and a movement of a target character in the image segment.

公开(公告)号：US20250036879A1

公开(公告)日：2025-01-30

申请号：US18740725

申请日：2024-06-12

Applicant: MASHANG CONSUMER FINANCE CO., LTD.

Inventor： Zhichao XIA ,  Xin WANG ,  Bing XIAO ,  Quan LU ,  Ning JIANG ,  Haiying WU

IPC: G06F40/30 ,  G06F40/284

Abstract: A method for sentiment analysis includes that: a feature sequence corresponding to text is obtained, the feature sequence includes encoded features; each encoded feature in the feature sequence is processed by using an attention mechanism, to obtain an attention feature of the text; the attention feature is transferred to a spatial transform feature of the text; an entity attribute of the spatial transform feature is recognized, and sentiment mapping is performed based on the spatial transform feature to obtain a sentiment polarity of the entity attribute.

公开(公告)号：US20240203081A1

公开(公告)日：2024-06-20

申请号：US18587550

申请日：2024-02-26

Applicant: MaShang Consumer Finance Co., Ltd.

Inventor： Yongchun LV ,  Hui ZHU ,  Xunyi ZHOU ,  Ning JIANG ,  Haiying WU

IPC: G06V10/22 ,  G06V10/77 ,  G06V10/771 ,  G06V10/774 ,  G06V10/82

CPC classification number: G06V10/22 ,  G06V10/771 ,  G06V10/7715 ,  G06V10/774 ,  G06V10/82

Abstract: Embodiments of the present application provide a processing method for an image recognition model and a related product. The processing method for the image recognition model includes: obtaining an image sample; determining a target object positioning box of the image sample, where the target object positioning box covers an area where an effective feature of the image sample is located; adjusting a pixel value of at least one pixel within the target object positioning box to obtain a preprocessed image sample; and training a network model based on the preprocessed image sample to obtain the image recognition model. The present application improves the concentration of the effective feature in the sample image and thus improves the regularization effect and reduces overfitting of the network model.