Self-supervised learning method and self-supervised learning apparatus

By monitoring and adjusting the mean distance of features in augmented images, the self-supervised learning method addresses the semantic bias problem in augmented images and improves the performance of computer vision tasks.

CN114092356BActive Publication Date: 2026-07-17JD DIGITS HAIYI INFORMATION TECHNOLOGY CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
JD DIGITS HAIYI INFORMATION TECHNOLOGY CO LTD
Filing Date
2021-11-26
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

In self-supervised learning, randomly generated augmented images may deviate from the semantic invariance assumption, leading to a decline in the performance of downstream computer vision tasks. Existing methods struggle to balance the variance and bias of data augmentation.

Method used

By monitoring the distance metric between each augmented image and the feature mean of each original image, weights are adjusted to suppress semantic bias. Self-supervised learning is then performed using a weighted cost function to ensure the diversity and consistency of the augmented images.

Benefits of technology

It improves the performance of self-supervised learning models on downstream computer vision tasks, effectively suppresses noise in augmented images, and enhances learning effectiveness.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114092356B_ABST
    Figure CN114092356B_ABST
Patent Text Reader

Abstract

The disclosure provides a self-supervised learning method and a self-supervised learning device, and relates to the field of computers. The disclosure can effectively monitor the augmented picture samples with semantic deviation through the distance measurement information of the feature of each augmented picture of each original picture to the feature mean of all augmented pictures of the original picture, and the training noise caused by the augmented picture samples with semantic deviation can be effectively inhibited by reducing the corresponding weight, thereby balancing the variance and deviation of the data augmentation distribution, and improving the performance of the learning model in downstream computer vision tasks.
Need to check novelty before this filing date? Find Prior Art