Regularized Distribution Matching Distillation for One-step Unpaired Image-to-Image Translation

?

Regularized Distribution Matching Distillation for One-step Unpaired Image-to-Image Translation

P. 1–21.

Diffusion distillation methods aim to compress the diffusion models into efficient one-step generators while trying to preserve quality. Among them, Distribution Matching Distillation (DMD) offers a suitable framework for training general-form onestep generators, applicable beyond unconditional generation. In this work, we introduce its modif ication, called Regularized Distribution Matching Distillation, applicable to unpaired image-toimage problems. We demonstrate its empirical performance in application to several translation tasks, including 2D examples and I2I between different image datasets, where it performs on par or better than multi-step diffusion baselines.

In book

ICML 2024 Workshop on Structured Probabilistic Inference & Generative Modeling

OpenReview, 2024.

Rethinking Global Text Conditioning in Diffusion Transformers

Nikita Starodubcev, Pakhomov D., Wu Z. et al., , in: The Fourteenth International Conference on Learning Representations (ICLR 2026).: ICLR, 2026.

Diffusion transformers typically incorporate textual information via (i) attention layers and (ii) a modulation mechanism using a pooled text embedding. Nevertheless, recent approaches discard modulation-based text conditioning and rely exclusively on attention. In this paper, we address whether modulation-based text conditioning is necessary and whether it can provide any performance advantage. Our analysis shows that, ...

Added: February 17, 2026

Диффузионные модели для генерации синтетических табличных данных

Телешева Э. Д., Hushchyn M., Доклады Российской академии наук. Математика, информатика, процессы управления (ранее - Доклады Академии Наук. Математика) 2025 Т. 527 № S С. 388–399

he problem of generating high-quality synthetic data is crucial for many data science tasks. A generated dataset can cut the costs on the augmentation of the existing data with additional instances, for example, in physics, or help with its privacy protection, for instance, in banking. However, generating a tabular dataset is challenging, as the data ...

Added: February 12, 2026

Перераспределение примеси на границе поликремний — кремний при моделировании диффузии из поликремния в TCAD при различных моделях диффузии

Федотов А. В., Sambursky L. M., Наноиндустрия 2025 Т. 18 № S11-2(135) С. 859–861

The parameters of the polysilicon—silicon interface in the simulation of thermal processes significantly depend on the choice of the diffusion model available in TCAD. In this paper, based on experimental data, a comparison of the diffusion models available in TCAD in application to these materials is carried out and recommendations for choosing a model are ...

Added: November 5, 2025

Generalization error bound for denoising score matching under relaxed manifold assumption

Yakovlev K., Puchkin N., , in: Proceedings of Machine Learning Research Vol. 291: The Thirty Eighth Annual Conference on Learning Theory, 30-4 July 2025, Lyon, France.: PMLR, 2025. P. 5824–5891.

We examine theoretical properties of the denoising score matching estimate. We model the density of observations with a nonparametric Gaussian mixture. We significantly relax the standard manifold assumption allowing the samples step away from the manifold. At the same time, we are still able to leverage a nice distribution structure. We derive non-asymptotic bounds on ...

Added: October 7, 2025

MaterialFusion: High-Quality, Zero-Shot, and Controllable Material Transfer with Diffusion Models

Garifullin K., Nikolaev Maxim, Kuznetsov A. et al., / Cornell University. Серия Computer Science "arxiv.org". 2025.

Manipulating the material appearance of objects in images is critical for applications like augmented reality, virtual prototyping, and digital content creation. We present MaterialFusion, a novel framework for high-quality material transfer that allows users to adjust the degree of material application, achieving an optimal balance between new material properties and the object’s original features. MaterialFusion ...

Added: February 24, 2025

MMCP Team at ImageCLEFmed MEDVQA-GI 2024 Task: Diffusion Models for Text-to-Image Generation of Colonoscopy Images

Chaichuk M., Tutubalina E., Transactions of the Association for Computational Linguistics 2024

This paper introduces models developed for the ImageCLEFmed 2024 MEDVQA-GI task, aimed at leveraging text-to-image generative models to create a comprehensive dataset of artificial colonoscopy images from textual prompts. The task’s complexity arises from the novel and relatively uncharted nature of the provided training dataset, its limited size, and the specificity required in the generated ...

Added: December 13, 2024

Group and Shuffle: Efficient Structured Orthogonal Parametrization

Gorbunov M., Yudin N., Soboleva V. et al., , in: 38th Conference on Neural Information Processing Systems (NeurIPS 2024).: [б.и.], 2024. P. 68713–68739.

Added: November 26, 2024

Guide-and-Rescale: Self-guidance Mechanism for Effective Tuning-Free Real Image Editing

Titov V., Khalmatova M., Ivanova A. A. et al., , in: 18th European Conference, Milan, Italy, September 29–October 4, 2024, Proceedings, Part LXXI. Computer Vision – ECCV 2024. (LNCS, volume 15129).: Springer, 2024. P. 235–251.

Despite recent advances in large-scale text-to-image generative models, manipulating real images with these models remains a challenging problem. The main limitations of existing editing methods are that they either fail to perform with consistent quality on a wide range of image edits or require time-consuming hyperparameter tuning or fine-tuning of the diffusion model to preserve ...

Added: November 26, 2024

Star-Shaped Denoising Diffusion Probabilistic Models

Andrey Okhotin, Dmitry Molchanov, Arkhipkin V. et al., , in: Advances in Neural Information Processing Systems 36 (NeurIPS 2023).: Curran Associates, Inc., 2023. P. 10038–10067.

Added: February 15, 2024