CORAL: Correspondence Alignment for Improved Virtual Try-On
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866915808077152256 |
|---|---|
| author | Kim, Jiyoung Shin, Youngjin Jin, Siyoon Chung, Dahyun Nam, Jisu Kim, Tongmin Park, Jongjae Kang, Hyeonwoo Kim, Seungryong |
| author_facet | Kim, Jiyoung Shin, Youngjin Jin, Siyoon Chung, Dahyun Nam, Jisu Kim, Tongmin Park, Jongjae Kang, Hyeonwoo Kim, Seungryong |
| contents | Existing methods for Virtual Try-On (VTON) often struggle to preserve fine garment details, especially in unpaired settings where accurate person-garment correspondence is required. These methods do not explicitly enforce person-garment alignment and fail to explain how correspondence emerges within Diffusion Transformers (DiTs). In this paper, we first analyze full 3D attention in DiT-based architecture and reveal that the person-garment correspondence critically depends on precise person-garment query-key matching within the full 3D attention. Building on this insight, we then introduce CORrespondence ALignment (CORAL), a DiT-based framework that explicitly aligns query-key matching with robust external correspondences. CORAL integrates two complementary components: a correspondence distillation loss that aligns reliable matches with person-garment attention, and an entropy minimization loss that sharpens the attention distribution. We further propose a VLM-based evaluation protocol to better reflect human preference. CORAL consistently improves over the baseline, enhancing both global shape transfer and local detail preservation. Extensive ablations validate our design choices. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_17636 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | CORAL: Correspondence Alignment for Improved Virtual Try-On Kim, Jiyoung Shin, Youngjin Jin, Siyoon Chung, Dahyun Nam, Jisu Kim, Tongmin Park, Jongjae Kang, Hyeonwoo Kim, Seungryong Computer Vision and Pattern Recognition Existing methods for Virtual Try-On (VTON) often struggle to preserve fine garment details, especially in unpaired settings where accurate person-garment correspondence is required. These methods do not explicitly enforce person-garment alignment and fail to explain how correspondence emerges within Diffusion Transformers (DiTs). In this paper, we first analyze full 3D attention in DiT-based architecture and reveal that the person-garment correspondence critically depends on precise person-garment query-key matching within the full 3D attention. Building on this insight, we then introduce CORrespondence ALignment (CORAL), a DiT-based framework that explicitly aligns query-key matching with robust external correspondences. CORAL integrates two complementary components: a correspondence distillation loss that aligns reliable matches with person-garment attention, and an entropy minimization loss that sharpens the attention distribution. We further propose a VLM-based evaluation protocol to better reflect human preference. CORAL consistently improves over the baseline, enhancing both global shape transfer and local detail preservation. Extensive ablations validate our design choices. |
| title | CORAL: Correspondence Alignment for Improved Virtual Try-On |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2602.17636 |