Passive domain adaptive target detection method based on pseudo tag confidence feedback and electronic equipment

By introducing a pseudo-label confidence feedback mechanism into the Mean-Teacher architecture, the EMA decay rate and pseudo-label screening threshold are dynamically adjusted, which solves the problem of pseudo-label noise accumulation in passive domain adaptive tasks, realizes stable model transfer and efficient training in the target domain, and improves detection performance.

CN122090211APending Publication Date: 2026-05-26SHANGHAI UNIV
0 Cites 0 Cited by

Patent Information

Application Number
CN202610522004.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-04-20
Publication Date
2026-05-26

AI Technical Summary

Technical Problem

The fixed EMA strategy in the existing Mean-Teacher architecture leads to the accumulation of pseudo-label noise, which causes the model to degrade detection performance in the target domain, and lacks effective pseudo-label quality control in passive domain adaptation tasks.

Method used

By introducing a pseudo-label confidence feedback mechanism, dynamically adjusting the pseudo-label screening threshold and EMA decay rate, an adaptive mean teacher update strategy is constructed. The confidence signal is used to optimize the weight update of the teacher model, suppress the accumulation of pseudo-label noise, and improve the model's transfer stability in the target domain.

Benefits of technology

It effectively suppressed the accumulation of false label noise, improved the training stability and detection accuracy of the model in the target domain, enhanced cross-domain transfer efficiency, reduced training costs, and strengthened the model's generalization ability.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention relates to the field of computer vision, in particular to a passive domain adaptive target detection method based on pseudo tag confidence feedback and electronic equipment, and the method comprises the following steps: S1, obtaining a pre-trained source domain model and unlabeled target domain data; s2, constructing a mean value teacher self-training architecture; s3, generating a pseudo tag and calculating a confidence feedback signal; s4, a self-adaptive dynamic mean value teacher updating strategy is carried out; according to the method, a pseudo label confidence feedback mechanism is introduced, so that the model has self-adaptive adjustment capability, and the adverse effect of low-quality pseudo labels on model training can be effectively inhibited; the attenuation rate can be dynamically reduced in a high-quality pseudo label scene, and the cross-domain migration efficiency and the detection precision are improved; according to the method, the problem of confirmation deviation in a passive domain adaptive task is effectively relieved, the training stability, noise immunity and cross-domain target detection performance of the model on a target domain are remarkably improved, and the method has good application value and popularization prospects.
Need to check novelty before this filing date? Find Prior Art