A data enhancement method, device, equipment and storage medium

By dividing comprehensive modal data into single-modal datasets and merging them to determine candidate data sets based on similarity, the problems of low accuracy and efficiency in the construction of multimodal datasets are solved, and efficient data augmentation and model training are achieved.

CN115705403BActive Publication Date: 2025-12-12TENCENT TECHNOLOGY (SHENZHEN) CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202110913888.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-08-10
Publication Date
2025-12-12
Estimated Expiration
2041-08-10

Smart Images

  • Figure CN115705403B_ABST
    Figure CN115705403B_ABST
Patent Text Reader

Abstract

Embodiments of the present application provide a data enhancement method and device, equipment and storage medium, relating to the technical field of artificial intelligence. In the method, a first comprehensive modal data set is divided into at least two single modal data sets of different modalities, each first comprehensive modal data including at least two single modal data of different modalities. Then, for each first comprehensive modal data, the similarity of each target single modal data in the first comprehensive modal data with each single modal data in the corresponding single modal data set is determined. Based on the obtained similarities, the candidate single modal data set corresponding to each target single modal data is determined from the single modal data set corresponding to each target single modal data. The obtained candidate single modal data sets are shared to determine new comprehensive modal data, thereby achieving data enhancement of a small amount of existing comprehensive modal data and improving the efficiency and accuracy of expanding comprehensive modal data.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Neural network model training method and device, computer equipment and storage medium

    CN109840588A

  • Data expansion method and device

    CN111159150A