Semi-Supervised Conditional Diffusion via Label Augmentation
Jin Su, Yuan Gao, Yong Zhou, Jian Huang
stat.ML, cs.LG
Submitted: 2026-07-18
Comments: 34 pages, 7 figures
License: http://creativecommons.org/licenses/by/4.0/
The gist: Conditional diffusion models have become a powerful and flexible framework for learning complex conditional distributions from labeled data.
Terminology
Abstract
Conditional diffusion models have become a powerful and flexible framework for learning complex conditional distributions from labeled data. In practice, however, acquiring high-quality labels is costly and time-consuming, leaving large volumes of unlabeled data unused. To address this, we introduce label-augmented conditional diffusion (LACD), a simple and effective approach that incorporates unlabeled examples by assigning them a designated trivial label and performing joint denoising score matching over the augmented dataset. We provide sufficient conditions guaranteeing population-level identifiability of the target conditional distribution under this scheme. Moreover, we establish rigorous statistical guarantees: when sufficiently many unlabeled samples are available, the sampling distribution produced by LACD converges strictly faster than the purely supervised estimator in total variation distance, and at least as fast in Wasserstein-1 distance. Extensive experiments on synthetic, image, and tabular benchmarks corroborate our theory and show substantial gains in sample efficiency and generative performance compared with the purely supervised estimator.
Sources
- Enhancing Causal Effect Estimation with Diffusion-Generated Data
- Unveil Conditional Diffusion Models with Classifier-free Guidance: A Sharp Statistical Theory
- Few-Shot Diffusion Models
- Classifier-Free Diffusion Guidance
- Score-Based Generative Modeling through Stochastic Differential Equations
- Enhancing Accuracy in Generative Models via Knowledge Transfer
- Information-acquiring von Neumann architecture of a computer: A theoretical design
- Counterfactual Generative Modeling with Variational Causal Inference
Related papers
- Behavior of prediction performance metrics with rare events
- Optimal Estimation of Generic Dynamics by Path-Dependent Neural Jump ODEs
- A Posterior-Dynamics Framework for Imaging Inverse Problems with Pretrained Diffusion Priors
- Scalable Posterior Uncertainty for Flexible Density-Based Clustering
- A Transport Geometry of Benign Overfitting
- Efficient quadratic entropy with distance sketches