发表机构
University of Toyama; Hokkaido University; University of Fukui（富山大学; 北海道大学; 福井大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对数据集蒸馏中基于扩散方法的局限性，提出显著性驱动蒸馏框架，通过两步构建类判别潜在原型，增强代表性与泛化能力，实验证明其性能优于强基线。
AI 中文摘要
数据集蒸馏旨在合成紧凑数据集，在大幅降低计算和存储成本的同时接近全数据训练的性能。然而，基于扩散的蒸馏方法难以保持结构一致性和泛化能力，尤其在视觉复杂领域。问题源于潜在原型与类判别区域对齐不佳且受无关背景影响。为此，我们提出显著性驱动的蒸馏框架，分两步构建类判别潜在原型以增强代表性和泛化能力。实验表明该框架性能优于强基线。代码将发布。
英文摘要
Dataset distillation aims to synthesize compact datasets that can approximate the performance of full-data training while significantly reducing computational and storage costs. However, diffusion-based distillation methods often struggle to preserve structural coherence and generalization, especially in visually complex domains. This issue often stems from latent prototypes that are weakly aligned with class-discriminative regions and contaminated by irrelevant background, thereby degrading generation quality and generalization. To address this limitation, we propose a saliency-driven distillation framework that constructs class-discriminative latent prototypes to enhance representativeness and generalization. The framework proceeds in two stages: (1) ensemble Grad-CAM++ saliency is used to construct prototypes emphasizing class-discriminative regions, and (2) hard-prototype refinement is then applied to construct challenging yet class-consistent prototypes, thereby enhancing discriminability and diversity. Importantly, the diffusion backbones (e.g., LDM and DiT) remain frozen; only lightweight classifiers used for saliency extraction are trained. Extensive experiments across multiple benchmarks demonstrate consistent performance improvements over strong baselines. Code will be released.
