公开(公告)号：US20240354582A1

公开(公告)日：2024-10-24

申请号：US18441076

申请日：2024-02-14

Applicant: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE

Inventor： Seon-Tae KIM ,  Joon-Sun AUH ,  Chang-Sik CHO

IPC: G06N3/0895

CPC classification number: G06N3/0895

Abstract: Disclosed herein are a method for contrastive learning of a neural network for automatic data labeling and an apparatus for the same. The method includes generating transformed data samples for unlabeled input data samples corresponding to a batch size, detecting a positive sample (positive data) in the transformed data samples in consideration of the similarity between a single data sample for contrastive learning, among the input data samples, and each of the transformed data samples, and performing contrastive learning of a neural network based on the loss value of the positive sample.