Active domain adaptation semantic segmentation method, system, device and storage medium

By employing superpixel-level annotation and a two-stage selection strategy, combined with cross-domain hybridization and pseudo-label consistency techniques, the problem of insufficient annotation efficiency and quality in existing technologies is solved, achieving efficient domain-adaptive semantic segmentation.

CN119131393BActive Publication Date: 2026-06-23UNIV OF SCI & TECH OF CHINA
6 Cites 0 Cited by

Patent Information

Application Number
CN202411254918.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-09
Publication Date
2026-06-23
Estimated Expiration
2044-09-09

AI Technical Summary

Technical Problem

In existing active domain-adaptive semantic segmentation methods, annotation efficiency and annotation quality need to be improved, and the domain adaptation performance has not been fully resolved.

Method used

A superpixel-level annotation method is adopted. The entropy map of the target domain image is extracted through the domain adaptation model. The superpixel extraction network is combined to divide the superpixels into high and low uncertainty. Two-stage annotation and training are carried out. High uncertainty superpixels with large feature differences are selected for annotation. The model is trained by combining cross-domain mixing and pseudo-label consistency techniques.

Benefits of technology

It improves annotation efficiency and quality, ensures the performance of domain-adaptive semantic segmentation, reduces annotation costs, and enhances the model's adaptability in the target domain.

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

Abstract

The application discloses an active domain adaptation semantic segmentation method, system, device and storage medium, selects and labels taking superpixels as units, which is different from image-level and pixel-level labeling methods, and the superpixel-level labeling greatly improves labeling efficiency by only assigning a semantic category to each superpixel; in addition, different from the existing scheme based on the uncertainty selection labeling strategy, the application focuses on difficult example samples in the domain adaptation scene, and proposes a selection strategy based on domain information quantity to label superpixels most valuable for domain adaptation learning; by adopting the superpixel-level labeling method and the selection strategy based on the domain information quantity, the application greatly reduces the labeling cost while improving the labeling quality, and guarantees the performance of the domain adaptation semantic segmentation.
Need to check novelty before this filing date? Find Prior Art