Label-Retrieval-Augmented Diffusion Models for Learning from Noisy Labels
Learning from noisy labels is an important and long-standing problem in machine learning for real applications. One of the main research lines focuses on learning a label corrector to purify potential noisy labels. However, these methods typically rely on strict assumptions and are limited to certain types of label noise. In this paper, we reformulate the label-noise problem from a generative-model perspective, $\textit{i.e.}$, labels are generated by gradually refining an initial random guess. This new perspective immediately enables existing powerful diffusion models to seamlessly learn the stochastic generative process. Once the generative uncertainty is modeled, we can perform classification inference using maximum likelihood estimation of labels. To mitigate the impact of noisy labels, we propose the $\textbf{L}$abel-$\textbf{R}$etrieval-$\textbf{A}$ugmented (LRA) diffusion model, which leverages neighbor consistency to effectively construct pseudo-clean labels for diffusion training. Our model is flexible and general, allowing easy incorporation of different types of conditional information, $\textit{e.g.}$, use of pre-trained models, to further boost model performance. Extensive experiments are conducted for evaluation. Our model achieves new state-of-the-art (SOTA) results on all the standard real-world benchmark datasets. Remarkably, by incorporating conditional information from the powerful CLIP model, our method can boost the current SOTA accuracy by 10-20 absolute points in many cases.
PDF Abstract NeurIPS 2023 PDF NeurIPS 2023 Abstract
CIFAR-10
CIFAR-100
Food-101
Clothing1M
WebVision
Results from the Paper
Ranked #1 on
Image Classification
on Food-101N
(using extra training data)
| Task | Dataset | Model | Metric Name | Metric Value | Global Rank | Uses Extra Training Data |
Benchmark |
|---|---|---|---|---|---|---|---|
| Image Classification | Clothing1M | LRA-diffusion (CC) | Accuracy | 75.7% | # 1 | ||
| Image Classification | Food-101N | LRA-diffusion (CLIP ViT) | Accuracy | 93.42 | # 1 | ||
| Image Classification | mini WebVision 1.0 | LRA-diffusion (CLIP ViT) | Top-1 Accuracy | 84.16 | # 1 | ||
| ImageNet Top-1 Accuracy | 82.56 | # 1 |
