Diffusion Dataset Condensation: Training Your Diffusion Model Faster with Less Data

• URL: http://arxiv.org/abs/2507.05914v2
• Date: Sat, 12 Jul 2025 09:02:47 GMT
• Title: Diffusion Dataset Condensation: Training Your Diffusion Model Faster with Less Data
• Authors: Rui Huang, Shitong Shao, Zikai Zhou, Pukun Zhao, Hangyu Guo, Tian Ye, Lichen Bai, Shuo Yang, Zeke Xie,
• Abstract summary: We study diffusion dataset condensation as a new and challenging problem setting.<n>The goal is to construct a "synthetic" sub-dataset with significantly fewer samples than the original dataset.<n>Our framework enables significantly faster diffusion model training with dramatically fewer data, while preserving high visual quality.
• Score: 18.21207020440108
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Diffusion models have achieved remarkable success in various generative tasks, but training them remains highly resource-intensive, often requiring millions of images and many days of GPU computation. From a data-centric perspective addressing this limitation, we study diffusion dataset condensation as a new and challenging problem setting. The goal is to construct a "synthetic" sub-dataset with significantly fewer samples than the original dataset, enabling high-quality diffusion model training with greatly reduced cost. To the best of our knowledge, we are the first to formally investigate dataset condensation for diffusion models, whereas prior work focused on training discriminative models. To tackle this new challenge, we propose a novel Diffusion Dataset Condensation (D2C) framework, which consists of two phases: Select and Attach. The Select phase identifies a compact and diverse subset using a diffusion difficulty score and interval sampling. The Attach phase enhances the selected subset by attaching rich semantic and visual representations to strengthen the conditional signals. Extensive experiments across various dataset sizes, model architectures, and resolutions show that our D2C framework enables significantly faster diffusion model training with dramatically fewer data, while preserving high visual quality. Notably, for the SiT-XL/2 architecture, D2C achieves a 100x training speed-up, reaching a FID score of 4.3 in just 40k steps using only 0.8% of the training data.

Related papers

• Diffusion-Augmented Coreset Expansion for Scalable Dataset Distillation [18.474302012851087]
  We propose a two-stage solution for dataset distillation.<n>First, we compress the dataset by selecting only the most informative patches to form a coreset.<n>Next, we leverage a generative foundation model to dynamically expand this compressed set in real-time.<n>We demonstrate a significant improvement of over 10% compared to the state-of-the-art on several large-scale dataset distillation benchmarks.
  arXiv  Detail & Related papers  (2024-12-05T23:40:27Z)
• Constrained Diffusion Models via Dual Training [80.03953599062365]
  Diffusion processes are prone to generating samples that reflect biases in a training dataset. We develop constrained diffusion models by imposing diffusion constraints based on desired distributions. We show that our constrained diffusion models generate new data from a mixture data distribution that achieves the optimal trade-off among objective and constraints.
  arXiv  Detail & Related papers  (2024-08-27T14:25:42Z)
• Gradient Inversion of Federated Diffusion Models [4.1355611383748005]
  Diffusion models are becoming defector generative models, which generate exceptionally high-resolution image data. In this paper, we study the privacy risk of gradient inversion attacks. We propose a triple-optimization GIDM+ that coordinates the optimization of the unknown data.
  arXiv  Detail & Related papers  (2024-05-30T18:00:03Z)
• Consistent Diffusion Meets Tweedie: Training Exact Ambient Diffusion Models with Noisy Data [74.2507346810066]
  Ambient diffusion is a recently proposed framework for training diffusion models using corrupted data. We present the first framework for training diffusion models that provably sample from the uncorrupted distribution given only noisy training data.
  arXiv  Detail & Related papers  (2024-03-20T14:22:12Z)
• Distribution-Aware Data Expansion with Diffusion Models [55.979857976023695]
  We propose DistDiff, a training-free data expansion framework based on the distribution-aware diffusion model. DistDiff consistently enhances accuracy across a diverse range of datasets compared to models trained solely on original data.
  arXiv  Detail & Related papers  (2024-03-11T14:07:53Z)
• DiffusionEngine: Diffusion Model is Scalable Data Engine for Object Detection [41.436817746749384]
  Diffusion Model is a scalable data engine for object detection. DiffusionEngine (DE) provides high-quality detection-oriented training pairs in a single stage.
  arXiv  Detail & Related papers  (2023-09-07T17:55:01Z)
• Improved Distribution Matching for Dataset Condensation [91.55972945798531]
  We propose a novel dataset condensation method based on distribution matching. Our simple yet effective method outperforms most previous optimization-oriented methods with much fewer computational resources.
  arXiv  Detail & Related papers  (2023-07-19T04:07:33Z)
• BOOT: Data-free Distillation of Denoising Diffusion Models with Bootstrapping [64.54271680071373]
  Diffusion models have demonstrated excellent potential for generating diverse images. Knowledge distillation has been recently proposed as a remedy that can reduce the number of inference steps to one or a few. We present a novel technique called BOOT, that overcomes limitations with an efficient data-free distillation algorithm.
  arXiv  Detail & Related papers  (2023-06-08T20:30:55Z)
• DiffusionSeg: Adapting Diffusion Towards Unsupervised Object Discovery [20.787180028571694]
  DiffusionSeg is a synthesis-exploitation framework containing two-stage strategies. We synthesize abundant images, and propose a novel training-free AttentionCut to obtain masks in the first stage. In the second exploitation stage, to bridge the structural gap, we use the inversion technique, to map the given image back to diffusion features.
  arXiv  Detail & Related papers  (2023-03-17T07:47:55Z)
• Accelerating Dataset Distillation via Model Augmentation [41.3027484667024]
  We propose two model augmentation techniques, i.e. using early-stage models and parameter parameters to learn an informative synthetic set with significantly reduced training cost. Our method achieves up to 20x speedup and comparable performance on par with state-of-the-art methods.
  arXiv  Detail & Related papers  (2022-12-12T07:36:05Z)
• On Distillation of Guided Diffusion Models [94.95228078141626]
  We propose an approach to distilling classifier-free guided diffusion models into models that are fast to sample from. For standard diffusion models trained on the pixelspace, our approach is able to generate images visually comparable to that of the original model. For diffusion models trained on the latent-space (e.g., Stable Diffusion), our approach is able to generate high-fidelity images using as few as 1 to 4 denoising steps.
  arXiv  Detail & Related papers  (2022-10-06T18:03:56Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.