Diff-CA: Separating Common and Salient Factors with Diffusion Models

• URL: http://arxiv.org/abs/2606.06120v1
• Date: Thu, 04 Jun 2026 13:07:18 GMT
• Title: Diff-CA: Separating Common and Salient Factors with Diffusion Models
• Abstract summary: We propose a novel conditioning framework for diffusion models that enables contrastive decomposition without compromising generation quality.<n>We prove that the additive contrastive factorization, commonly assumed in prior work, is identifiable under mild conditions.<n>This factorization enables targeted operations by swapping or interpolating only the salient factor.
• Score: 43.80684668106847
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Contrastive Analysis aims to separate factors that are common between two data distributions from those that are salient to only one of them. Existing contrastive methods are based on generative models (e.g., VAEs or GANs) that often suffer from limited reconstruction and image quality, which hampers effective latent factor separation and limits their applicability to high-fidelity image generation and edition. We propose a novel conditioning framework for diffusion models that enables contrastive decomposition without compromising generation quality. We first train a prompt-free, image-conditioned diffusion model, and then learn to decompose the conditioning into a common and a salient factor, using weak supervision. We prove that the additive contrastive factorization, commonly assumed in prior work, is identifiable under mild conditions. This factorization enables targeted operations by swapping or interpolating only the salient factor.

Related papers

• Interpolating Discrete Diffusion Models with Controllable Resampling [49.771004594828405]
  Masked diffusion models suffer from irreversible errors due to early unmasking.<n>Uniform diffusion models, despite enabling self-correction, often yield low-quality samples.<n>We introduce Interpolating Discrete Diffusion Model, that improves diffusion by reducing dependence on intermediate latent states.
  arXiv  Detail & Related papers  (2026-04-19T07:55:57Z)
• Solving Inverse Problems with FLAIR [68.87167940623318]
  We present FLAIR, a training-free variational framework that leverages flow-based generative models as prior for inverse problems.<n>Results on standard imaging benchmarks demonstrate that FLAIR consistently outperforms existing diffusion- and flow-based methods in terms of reconstruction quality and sample diversity.
  arXiv  Detail & Related papers  (2025-06-03T09:29:47Z)
• TRAIL: Transferable Robust Adversarial Images via Latent diffusion [35.54430200195499]
  Adversarial attacks present severe security risks to deep learning systems.<n> transferability across models remains limited due to distribution mismatches between generated adversarial features and real-world data.<n>We propose Transferable Robust Adrial Images via Latent Diffusion (TRAIL), a test-time adaptation framework.
  arXiv  Detail & Related papers  (2025-05-22T03:11:35Z)
• VENOM: Text-driven Unrestricted Adversarial Example Generation with Diffusion Models [26.513728933354958]
  VENOM is the first framework for high-quality unrestricted adversarial examples generation through diffusion models.<n>We introduce an adaptive adversarial guidance strategy with momentum, ensuring that the generated adversarial examples align with the distribution $p(x)$ of deceiving natural images.
  arXiv  Detail & Related papers  (2025-01-14T08:12:20Z)
• Distillation of Discrete Diffusion through Dimensional Correlations [21.078500510691747]
  "Mixture" models are capable of treating dimensional correlations while remaining scalable.<n>Loss functions enable the mixture models to distill such many-step conventional models into just a few steps by learning the dimensional correlations.<n>Results show the effectiveness of the proposed method in distilling pretrained discrete diffusion models across image and language domains.
  arXiv  Detail & Related papers  (2024-10-11T10:53:03Z)
• Membership Inference on Text-to-Image Diffusion Models via Conditional Likelihood Discrepancy [36.156856772794065]
  We propose a conditional overfitting phenomenon in text-to-image diffusion models. Our method significantly outperforms previous methods across various data and dataset scales.
  arXiv  Detail & Related papers  (2024-05-23T17:09:51Z)
• JoReS-Diff: Joint Retinex and Semantic Priors in Diffusion Model for Low-light Image Enhancement [69.6035373784027]
  Low-light image enhancement (LLIE) has achieved promising performance by employing conditional diffusion models. Previous methods may neglect the importance of a sufficient formulation of task-specific condition strategy. We propose JoReS-Diff, a novel approach that incorporates Retinex- and semantic-based priors as the additional pre-processing condition.
  arXiv  Detail & Related papers  (2023-12-20T08:05:57Z)
• Steered Diffusion: A Generalized Framework for Plug-and-Play Conditional Image Synthesis [62.07413805483241]
  Steered Diffusion is a framework for zero-shot conditional image generation using a diffusion model trained for unconditional generation. We present experiments using steered diffusion on several tasks including inpainting, colorization, text-guided semantic editing, and image super-resolution.
  arXiv  Detail & Related papers  (2023-09-30T02:03:22Z)
• Your Diffusion Model is Secretly a Zero-Shot Classifier [90.40799216880342]
  We show that density estimates from large-scale text-to-image diffusion models can be leveraged to perform zero-shot classification. Our generative approach to classification attains strong results on a variety of benchmarks. Our results are a step toward using generative over discriminative models for downstream tasks.
  arXiv  Detail & Related papers  (2023-03-28T17:59:56Z)
• SinDiffusion: Learning a Diffusion Model from a Single Natural Image [159.4285444680301]
  We present SinDiffusion, leveraging denoising diffusion models to capture internal distribution of patches from a single natural image. It is based on two core designs. First, SinDiffusion is trained with a single model at a single scale instead of multiple models with progressive growing of scales. Second, we identify that a patch-level receptive field of the diffusion network is crucial and effective for capturing the image's patch statistics.
  arXiv  Detail & Related papers  (2022-11-22T18:00:03Z)

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.