mlsanityData to 14 Oct 2026

MAST: Mask-Guided Attention Control for Training-Free Regional-Multi Style Transfer

arXiv 2604.12281 ↗First posted 14 Apr 2026cs.CV cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Style transfer applies the appearance of a reference image to a content image while preserving its spatial structure. Recent diffusion-based methods achieve strong stylization but typically assume a single global style. We instead consider regional-multi style transfer, which assigns multiple references to user-specified regions of a content image. Extending them to this setting reveals two coupled shared-attention issues: ambiguous mass allocation among content and style partitions and degraded selectivity as more styles are jointly normalized, while style aggregation at the attention output further suppresses fine details. We propose MAST (Mask-Guided Attention Control for Training-Free Regional-Multi Style Transfer), a unified attention-control framework for frozen diffusion models. Logit-level Attention Mass Allocation enforces mask-derived partition masses, Sharpness-aware Temperature Scaling adaptively restores selectivity, and Discrepancy-aware Detail Injection recovers high-frequency content. MAST jointly processes all style--mask pairs in a single denoising pass without training, optimization, or post-hoc composition. Across two to five styles, MAST achieves the best average ArtFID, FID, and R-FID among all baselines, demonstrating regional style fidelity, content preservation, and scalability.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 29 of its 45 references that are in the map. First posted on arXiv 14 Apr 2026.

Builds on

29 of 45 references.

References · 29

  1. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  2. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  3. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
  4. 2020Denoising Diffusion Implicit Modelscited by 9,781
  5. 2017Arbitrary Style Transfer in Real-time with Adaptive Instance Normalizationcited by 2,010
  6. 2017Universal Style Transfer via Feature Transformscited by 354
  7. 2016A Learned Representation For Artistic Stylecited by 341
  8. 2022Inversion-Based Style Transfer with Diffusion Modelscited by 291
  9. 2023Fast Segment Anythingcited by 290
  10. 2023Style Injection in Diffusion: A Training-free Approach for Adapting Large-scale Diffusion Models for Style Transfercited by 196
  11. 2021StyTr2: Image Style Transfer with Transformerscited by 187
  12. 2023StyleDiffusion: Controllable Disentangled Style Transfer via Diffusion Modelscited by 155
17 more
  1. 2022ArtFID: Quantitative Evaluation of Neural Style Transfercited by 49
  2. 2024Softmax is not Enough (for Sharp Size Generalisation)cited by 46
  3. 2025StyleSSP: Sampling StartPoint Enhancement for Training-free Diffusion-based Method for Style Transfercited by 23
  4. 2023Any-to-Any Style Transfer: Making Picasso and Da Vinci Collaboratecited by 17
  5. 2022Fine-Grained Image Style Transfer with Visual Transformerscited by 11
  6. 2025AttenST: A Training-Free Attention-Driven Style Transfer Framework with Pre-Trained Diffusion Modelscited by 9
  7. 2025ConsisLoRA: Enhancing Content and Style Consistency for LoRA-based Style Transfercited by 9
  8. 2019Style Mixer: Semantic-aware Multi-Style Transfer Networkcited by 8
  9. 2022Controllable Style Transfer via Test-time Training of Implicit Neural Representationcited by 8
  10. 2024An Efficient and Harmonized Framework for Balanced Cross-Domain Feature Integrationcited by 7
  11. 2024DiffuseST: Unleashing the Capability of the Diffusion Model for Style Transfercited by 5
  12. 2026StyleGallery: Training-free and Semantic-aware Personalized Style Transfer from Arbitrary Image Referencescited by 2
  13. 2025Improving Masked Style Transfer using Blended Partial Convolutioncited by 1
  14. 2025Leveraging Diffusion Models for Stylization using Multiple Style Imagescited by 1
  15. 2025SCAdapter: Content-Style Disentanglement for Diffusion Style Transfercited by 1
  16. 2026CoCoDiff: Correspondence-Consistent Diffusion Model for Fine-grained Style Transfercited by 1
  17. 2026RegionRoute: Regional Style Transfer with Diffusion Modelcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026HAM: A Training-Free Style Transfer Approach via Heterogeneous Attention Modulation for Diffusion Models
  2. 2026TRACE: Time-Adaptive Residual Attention Control with Content-Style Decomposition for Training-Free Diffusion Style Transfer
  3. 2025ICAS: IP Adapter and ControlNet-based Attention Structure for Multi-Subject Style Transfer Optimization
  4. 2026A training-free framework for high-fidelity appearance transfer via diffusion transformers
  5. 2026Local Content-Style Control for Diffusion-based Image Stylization
  6. 2026Scheduled Style Injection: Expanding the Style-Content Pareto Frontier in Training-Free Diffusion-based Style Transfer
  7. 2023Z^*: Zero-shot Style Transfer via Attention Rearrangement
  8. 2025Training-Free Multi-Style Fusion Through Reference-Based Adaptive Modulation