Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
September 10, 2025Remote SensingOpen Access

Single-Domain Generalization via Multilevel Data Augmentation for SAR Target Recognition Training on Fully Simulated Data

View Full Paper
Ask AI
Bookmark
Share

Authors

WSWenyu ShuRZRonghui ZhanSCShiqi Chen

Discussion

Loading...

Member takes

Overview

Proposed method enhances SAR-ATR model accuracy against domain shift, implying improved generalization capabilities.

Key Points

  • The method achieves 96.76% accuracy on the SAMPLE dataset, significantly improving recognition performance.
  • Utilizing feature-level style augmentation, it enhances diversity through probabilistic mixing of instance-wise features.
  • Pixel-level augmentation enriches data distribution by injecting random Gaussian noise, improving pixel diversity.
  • Domain-adversarial training enforces learning of domain-invariant representations for better model adaptation.

Cite This Study

Shu et al. (2025) studied this question.

synapsesocial.com/papers/68c1d5ef54b1d3bfb60f8d20https://doi.org/10.3390/rs17172966
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Synthetic SAR data domain randomization for unseen SAR ATR2024
  2. 2Air Target ISAR Recognition Based on Data Augmentation and Transfer Learning2026
  3. 3Class Ambiguity Regularized Adversarial Domain Adaptation for Cross-resolution SAR ATR2024
  4. 4Coarse-to-Fine Structure and Semantic Learning for Single-Sample SAR Image Generation2024 · 2 citations
  5. 5Attribute Feature Perturbation-Based Augmentation of SAR Target Data2024