.png){kind=icon}
返回
GIDD: Gradient inversion using diffusion model for denoising in federated learning
.png){kind=icon}
PRE
AI
DOI：10.1016/j.neucom.2026.132682.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
• Introduces GIDD, a diffusion-based GIA with staged reverse denoising coupled with gradient matching for stable, controllable reconstruction. • Provides theory-informed justification: under mild Lipschitz conditions, block-wise optimization admits approximately monotonic descent. • Proposes QRC-PSNR, a unified metric combining quality, computational efficiency, and hardware resource usage. • Evaluates on CelebA and ImageNet, including classical and hybrid defenses, demonstrating competitive overall performance and efficiency.
期刊
IF:
6.5
论文数:
2.5W
被引数:
6.5W
机构
暂无机构信息
引用论文
Diffusion model with disentangled modulations for sharpening multispectral and hyperspectral images具有解纠缠调制的扩散模型，用于锐化多光谱和高光谱图像.png)
INFORMATION FUSION
INFORMATION FUSIONIF15.5
PREAI
分享收藏 没有更多内容
没有更多内容