SAS-Net: Cross-Domain Image Registration as Inverse Rendering via Structure-Appearance Factorization
Fuente:
arXiv
Saved in:
| Main Author: | |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914392333877248 |
|---|---|
| author | Qin, Jiahao |
| author_facet | Qin, Jiahao |
| contents | Cross-domain image registration requires aligning images acquired under heterogeneous imaging physics, where the classical brightness constancy assumption is fundamentally violated. We formulate this problem through an image formation model I = R(s, a) + epsilon, where each observation is generated by a rendering function R acting on domain-invariant scene structure s and domain-specific appearance statistics a. Registration then reduces to an inverse rendering problem: given observations from two domains, recover the shared structure and re-render it under the target appearance to obtain the registered output. We instantiate this framework as SAS-Net (Scene-Appearance Separation Network), where instance normalization implements the structure-appearance decomposition and Adaptive Instance Normalization (AdaIN) realizes the differentiable forward renderer. A scene consistency loss enforces geometric correspondence in the factorized latent space. Experiments on EuroSAT-Reg-256 (satellite remote sensing) and FIRE-Reg-256 (retinal fundus) demonstrate state-of-the-art performance across heterogeneous imaging domains. SAS-Net (3.35M parameters) achieves 89 FPS on an RTX 5090 GPU. Code:
https://github.com/D-ST-Sword/SAS-Net. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_09050 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | SAS-Net: Cross-Domain Image Registration as Inverse Rendering via Structure-Appearance Factorization Qin, Jiahao Image and Video Processing Artificial Intelligence Computer Vision and Pattern Recognition 68U10, 92C55 I.4.3; I.2.6; J.3 Cross-domain image registration requires aligning images acquired under heterogeneous imaging physics, where the classical brightness constancy assumption is fundamentally violated. We formulate this problem through an image formation model I = R(s, a) + epsilon, where each observation is generated by a rendering function R acting on domain-invariant scene structure s and domain-specific appearance statistics a. Registration then reduces to an inverse rendering problem: given observations from two domains, recover the shared structure and re-render it under the target appearance to obtain the registered output. We instantiate this framework as SAS-Net (Scene-Appearance Separation Network), where instance normalization implements the structure-appearance decomposition and Adaptive Instance Normalization (AdaIN) realizes the differentiable forward renderer. A scene consistency loss enforces geometric correspondence in the factorized latent space. Experiments on EuroSAT-Reg-256 (satellite remote sensing) and FIRE-Reg-256 (retinal fundus) demonstrate state-of-the-art performance across heterogeneous imaging domains. SAS-Net (3.35M parameters) achieves 89 FPS on an RTX 5090 GPU. Code: https://github.com/D-ST-Sword/SAS-Net. |
| title | SAS-Net: Cross-Domain Image Registration as Inverse Rendering via Structure-Appearance Factorization |
| topic | Image and Video Processing Artificial Intelligence Computer Vision and Pattern Recognition 68U10, 92C55 I.4.3; I.2.6; J.3 |
| url | https://arxiv.org/abs/2602.09050 |