Distribution-Aware Data Expansion with Diffusion Models

TL;DR

DistDiff is a distribution-aware diffusion model framework that enhances dataset expansion by generating distribution-consistent samples, improving model accuracy and robustness without requiring additional training.

Contribution

The paper introduces DistDiff, a novel, training-free data expansion method using hierarchical diffusion models that maintains distribution fidelity and outperforms existing synthesis techniques.

Findings

DistDiff improves accuracy across multiple datasets.

It outperforms existing synthesis-based data augmentation methods.

The expanded datasets enhance robustness across different model architectures.

Abstract

The scale and quality of a dataset significantly impact the performance of deep models. However, acquiring large-scale annotated datasets is both a costly and time-consuming endeavor. To address this challenge, dataset expansion technologies aim to automatically augment datasets, unlocking the full potential of deep models. Current data expansion techniques include image transformation and image synthesis methods. Transformation-based methods introduce only local variations, leading to limited diversity. In contrast, synthesis-based methods generate entirely new content, greatly enhancing informativeness. However, existing synthesis methods carry the risk of distribution deviations, potentially degrading model performance with out-of-distribution samples. In this paper, we propose DistDiff, a training-free data expansion framework based on the distribution-aware diffusion model. DistDiff constructs hierarchical prototypes to approximate the real data distribution, optimizing latent data points within diffusion models with hierarchical energy guidance. We demonstrate its capability to generate distribution-consistent samples, significantly improving data expansion tasks. DistDiff consistently enhances accuracy across a diverse range of datasets compared to models trained solely on original data. Furthermore, our approach consistently outperforms existing synthesis-based techniques and demonstrates compatibility with widely adopted transformation-based augmentation methods. Additionally, the expanded dataset exhibits robustness across various architectural frameworks. Our code is available at https://github.com/haoweiz23/DistDiff