Enhancing Diffusion-based Dataset Distillation via Adversary-Guided Curriculum Sampling

TL;DR

This paper introduces Adversary-guided Curriculum Sampling (ACS), a novel method that improves diffusion-based dataset distillation by enhancing diversity and coverage, leading to better performance on large-scale datasets.

Contribution

The paper proposes ACS, which guides diffusion sampling with an adversarial loss to create more diverse and comprehensive distilled datasets, addressing redundancy issues in existing methods.

Findings

Achieves 4.1% improvement on Imagewoof

Achieves 2.1% improvement on ImageNet-1k

Enhances dataset diversity and coverage

Abstract

Dataset distillation aims to encapsulate the rich information contained in dataset into a compact distilled dataset but it faces performance degradation as the image-per-class (IPC) setting or image resolution grows larger. Recent advancements demonstrate that integrating diffusion generative models can effectively facilitate the compression of large-scale datasets while maintaining efficiency due to their superiority in matching data distribution and summarizing representative patterns. However, images sampled from diffusion models are always blamed for lack of diversity which may lead to information redundancy when multiple independent sampled images are aggregated as a distilled dataset. To address this issue, we propose Adversary-guided Curriculum Sampling (ACS), which partitions the distilled dataset into multiple curricula. For generating each curriculum, ACS guides diffusion sampling process by an adversarial loss to challenge a discriminator trained on sampled images, thus mitigating information overlap between curricula and fostering a more diverse distilled dataset. Additionally, as the discriminator evolves with the progression of curricula, ACS generates images from simpler to more complex, ensuring efficient and systematic coverage of target data informational spectrum. Extensive experiments demonstrate the effectiveness of ACS, which achieves substantial improvements of 4.1\% on Imagewoof and 2.1\% on ImageNet-1k over the state-of-the-art.