mlsanityData to 14 Oct 2026

Rethinking Garment Conditioning in Diffusion-based Virtual Try-On: Decouple, Don't Denoise

arXiv 2511.18775 ↗First posted 24 Nov 2025cs.CV cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Virtual Try-On (VTON) synthesizes realistic images of a person wearing a target garment, with broad applications in e-commerce and fashion. Diffusion-based dual-UNet methods achieve strong results but double the parameters by dedicating a separate network to garment conditioning. Spatial concatenation offers a simpler single-network alternative, yet both UNet- and DiT-based instantiations report that full fine-tuning is ineffective, and the community has settled for attention-only training. We ask: why does full fine-tuning fail, and can this be resolved? Through what is, to our knowledge, the first visualization study of dual-UNet reference network behavior, we identify a unifying insight: garment conditioning must be decoupled from the denoising process. Spatial concatenation violates this by embedding the garment within the denoising target, causing three conflicts: guidance leakage, gradient competition, and train-test discrepancy. We derive three design principles to restore this decoupling and implement them as a pure recipe atop a standard architecture with no modification. The resulting model, DeCo-VTON (860M params), achieves single-network state of the art, matching the dual-UNet state of the art at half the cost while being preferred in human evaluation.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 46 of its 59 references that are in the map. First posted on arXiv 24 Nov 2025.

Builds on

46 of 59 references.

References · 46

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  3. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  4. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  5. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
  6. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,974
  7. 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 6,013
  8. 2022Classifier-Free Diffusion Guidancecited by 5,847
  9. 2022Generative Adversarial Networkscited by 5,627
  10. 2023Adding Conditional Control to Text-to-Image Diffusion Modelscited by 5,503
  11. 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesiscited by 4,356
  12. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesiscited by 4,181
34 more
  1. 2022Elucidating the Design Space of Diffusion-Based Generative Modelscited by 3,045
  2. 2022An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversioncited by 2,245
  3. 2021SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equationscited by 1,937
  4. 2023IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image Diffusion Modelscited by 1,464
  5. 2018Demystifying MMD GANscited by 1,293
  6. 2018DensePose: Dense Human Pose Estimation In The Wildcited by 575
  7. 2021VITON-HD: High-Resolution Virtual Try-On via Misalignment-Aware Normalizationcited by 287
  8. 2017VITON: An Image-based Virtual Try-on Networkcited by 286
  9. 2024BrushNet: A Plug-and-Play Image Inpainting Model with Decomposed Dual-Branch Diffusioncited by 202
  10. 2018Toward Characteristic-Preserving Image-based Virtual Try-On Networkcited by 192
  11. 2019Self-Correction for Human Parsingcited by 180
  12. 2023StableVITON: Learning Semantic Correspondence with Latent Diffusion Model for Virtual Try-Oncited by 175
  13. 2022Dress Code: High-Resolution Multi-Category Virtual Try-Oncited by 162
  14. 2023LaDI-VTON: Latent Diffusion Textual-Inversion Enhanced Virtual Try-Oncited by 152
  15. 2024OOTDiffusion: Outfitting Fusion based Latent Diffusion for Controllable Virtual Try-oncited by 151
  16. 2020Towards Photo-Realistic Virtual Try-On by Adaptively GeneratingleftrightarrowPreserving Image Contentcited by 149
  17. 2023TryOnDiffusion: A Tale of Two UNetscited by 141
  18. 2022High-Resolution Virtual Try-On with Misalignment and Occlusion-Handled Conditionscited by 130
  19. 2023Understanding the Latent Space of Diffusion Models through the Lens of Riemannian Geometrycited by 125
  20. 2023Taming the Power of Diffusion Models for High-Quality Virtual Try-On with Appearance Flowcited by 124
  21. 2023GP-VTON: Towards General Purpose Virtual Try-on via Collaborative Local-Flow Global-Parsing Learningcited by 120
  22. 2017Look into Person: Self-supervised Structure-sensitive Learning and A New Benchmark for Human Parsingcited by 109
  23. 2024Improving Diffusion Models for Authentic Virtual Try-on in the Wildcited by 109
  24. 2024CatVTON: Concatenation Is All You Need for Virtual Try-On with Diffusion Modelscited by 102
  25. 2015Deep Human Parsing with Active Template Regressioncited by 51
  26. 2024CleanDIFT: Diffusion Features without Noisecited by 44
  27. 2021Towards Scalable Unpaired Virtual Try-On via Patch-Routed Spatially-Adaptive GANcited by 40
  28. 2022C-VTON: Context-Driven Image-Based Virtual Try-On Networkcited by 29
  29. 2024Learning Flow Fields in Attention for Controllable Person Image Generationcited by 27
  30. 2024PromptDresser: Improving the Quality and Controllability of Virtual Try-On via Generative Textual Prompt and Prompt-aware Maskcited by 17
  31. 2024D4-VTON: Dynamic Semantics Disentangling for Differential Diffusion based Virtual Try-Oncited by 14
  32. 2023DREAM: Diffusion Rectification and Estimation-Adaptive Modelscited by 12
  33. 2025Voost: A Unified and Scalable Diffusion Transformer for Bidirectional Virtual Try-On and Try-Offcited by 8
  34. 2025RefTon: Reference person shot assist virtual Try-oncited by 1

Built on by

Not cited yet.