VP-NTK: Exploring the Benefits of Visual Prompting in Differentially Private Data Synthesis
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866915206604521472 |
|---|---|
| author | Hsu, Chia-Yi Chen, Jia-You Tsai, Yu-Lin Lin, Chih-Hsun Chen, Pin-Yu Yu, Chia-Mu Huang, Chun-Ying |
| author_facet | Hsu, Chia-Yi Chen, Jia-You Tsai, Yu-Lin Lin, Chih-Hsun Chen, Pin-Yu Yu, Chia-Mu Huang, Chun-Ying |
| contents | Differentially private (DP) synthetic data has become the de facto standard for releasing sensitive data. However, many DP generative models suffer from the low utility of synthetic data, especially for high-resolution images. On the other hand, one of the emerging techniques in parameter efficient fine-tuning (PEFT) is visual prompting (VP), which allows well-trained existing models to be reused for the purpose of adapting to subsequent downstream tasks. In this work, we explore such a phenomenon in constructing captivating generative models with DP constraints. We show that VP in conjunction with DP-NTK, a DP generator that exploits the power of the neural tangent kernel (NTK) in training DP generative models, achieves a significant performance boost, particularly for high-resolution image datasets, with accuracy improving from 0.644$\pm$0.044 to 0.769. Lastly, we perform ablation studies on the effect of different parameters that influence the overall performance of VP-NTK. Our work demonstrates a promising step forward in improving the utility of DP synthetic data, particularly for high-resolution images. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_16195 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | VP-NTK: Exploring the Benefits of Visual Prompting in Differentially Private Data Synthesis Hsu, Chia-Yi Chen, Jia-You Tsai, Yu-Lin Lin, Chih-Hsun Chen, Pin-Yu Yu, Chia-Mu Huang, Chun-Ying Computer Vision and Pattern Recognition Machine Learning Differentially private (DP) synthetic data has become the de facto standard for releasing sensitive data. However, many DP generative models suffer from the low utility of synthetic data, especially for high-resolution images. On the other hand, one of the emerging techniques in parameter efficient fine-tuning (PEFT) is visual prompting (VP), which allows well-trained existing models to be reused for the purpose of adapting to subsequent downstream tasks. In this work, we explore such a phenomenon in constructing captivating generative models with DP constraints. We show that VP in conjunction with DP-NTK, a DP generator that exploits the power of the neural tangent kernel (NTK) in training DP generative models, achieves a significant performance boost, particularly for high-resolution image datasets, with accuracy improving from 0.644$\pm$0.044 to 0.769. Lastly, we perform ablation studies on the effect of different parameters that influence the overall performance of VP-NTK. Our work demonstrates a promising step forward in improving the utility of DP synthetic data, particularly for high-resolution images. |
| title | VP-NTK: Exploring the Benefits of Visual Prompting in Differentially Private Data Synthesis |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2503.16195 |