LPH-VTON: Resolving the Structure-Texture Dilemma of Virtual Try-On via Latent Process Handover
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911685769428992 |
|---|---|
| author | Liu, Yixin Qian, Baihong Jiang, Jinglin Wu, Jeffery Chen, Yan Wang, Wei Wang, Yida Yang, Lanqing Xue, Guangtao |
| author_facet | Liu, Yixin Qian, Baihong Jiang, Jinglin Wu, Jeffery Chen, Yan Wang, Wei Wang, Yida Yang, Lanqing Xue, Guangtao |
| contents | Virtual Try-On (VTON) aims to synthesize photorealistic images of garments precisely aligned with a person's body and pose. Current diffusion-based methods, however, face a fundamental trade-off between structural integrity and textural fidelity. In this paper, we formalize this challenge as a consequence of complementary inductive biases inherent in prevailing architectures: models heavily reliant on spatial constraints naturally favor geometric alignment but often suppress textures, whereas models dominated by unconstrained generative priors excel at vibrant detail rendering but are prone to structural drift. Based on this diagnosis, we propose LPH-VTON, a new synergistic framework that resolves this tension within a single, continuous denoising process. LPH-VTON strategically decomposes the generation, leveraging a structure-biased model to establish a geometrically consistent latent scaffold in the early stages, before handing over control to a texture-biased model for high-fidelity detail rendering. Extensive experiments validate our approach. Our model achieves a superior Pareto-optimal balance, establishing new benchmarks in perceptual faithfulness while maintaining highly competitive structural alignment across the standard dataset VITON-HD, proving the efficacy of temporal architectural decoupling. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_14874 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | LPH-VTON: Resolving the Structure-Texture Dilemma of Virtual Try-On via Latent Process Handover Liu, Yixin Qian, Baihong Jiang, Jinglin Wu, Jeffery Chen, Yan Wang, Wei Wang, Yida Yang, Lanqing Xue, Guangtao Computer Vision and Pattern Recognition Virtual Try-On (VTON) aims to synthesize photorealistic images of garments precisely aligned with a person's body and pose. Current diffusion-based methods, however, face a fundamental trade-off between structural integrity and textural fidelity. In this paper, we formalize this challenge as a consequence of complementary inductive biases inherent in prevailing architectures: models heavily reliant on spatial constraints naturally favor geometric alignment but often suppress textures, whereas models dominated by unconstrained generative priors excel at vibrant detail rendering but are prone to structural drift. Based on this diagnosis, we propose LPH-VTON, a new synergistic framework that resolves this tension within a single, continuous denoising process. LPH-VTON strategically decomposes the generation, leveraging a structure-biased model to establish a geometrically consistent latent scaffold in the early stages, before handing over control to a texture-biased model for high-fidelity detail rendering. Extensive experiments validate our approach. Our model achieves a superior Pareto-optimal balance, establishing new benchmarks in perceptual faithfulness while maintaining highly competitive structural alignment across the standard dataset VITON-HD, proving the efficacy of temporal architectural decoupling. |
| title | LPH-VTON: Resolving the Structure-Texture Dilemma of Virtual Try-On via Latent Process Handover |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2605.14874 |