High-Order Vehicular Pattern Learning and Privacy-Preserving and Unsupervised GAN for Privacy Protection Toward Vehicular Parts Detection
This paper introduces High-order Vehicular Pattern Learning (HVPL), a novel framework designed to enhance vehicular pattern detection while ensuring privacy protection, associated with authentication through the integration of Privacy-Preserving and Unsupervised GAN (PPUP-GAN). To preserve data priv...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11029001/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This paper introduces High-order Vehicular Pattern Learning (HVPL), a novel framework designed to enhance vehicular pattern detection while ensuring privacy protection, associated with authentication through the integration of Privacy-Preserving and Unsupervised GAN (PPUP-GAN). To preserve data privacy, we use the PPUP-GAN to create massive-scale vehicular data that well mimics real data. Afterward, HVPL improves upon traditional methods by constructing a large-scale hypergraph that models complex relationships among vehicular pattern features, uncovering high-order interactions that conventional models fail to detect. The model also uses a Gaussian Mixture Model (GMM)-based posterior probability to rank candidate patches for vehicular pattern detection, prioritizing high-risk areas and minimizing false positives. Experiments demonstrated that our method outperforms its counterparts by achieving higher precision and stability in vehicular pattern detection, with a notable improvement of up to 33.39% in Average Precision (AP) and 62.62% in AP_50. |
|---|---|
| ISSN: | 2169-3536 |