DNA-Prior: Unsupervised Denoise Anything via Dual-Domain Prior
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866915643212693504 |
|---|---|
| author | Cheng, Yanqi Cheng, Chun-Wun Denholm, Jim Lima, Thiago Montoya-Zegarra, Javier A. Goodwin, Richard Schönlieb, Carola-Bibiane Aviles-Rivero, Angelica I |
| author_facet | Cheng, Yanqi Cheng, Chun-Wun Denholm, Jim Lima, Thiago Montoya-Zegarra, Javier A. Goodwin, Richard Schönlieb, Carola-Bibiane Aviles-Rivero, Angelica I |
| contents | Medical imaging pipelines critically rely on robust denoising to stabilise downstream tasks such as segmentation and reconstruction. However, many existing denoisers depend on large annotated datasets or supervised learning, which restricts their usability in clinical environments with heterogeneous modalities and limited ground-truth data. To address this limitation, we introduce DNA-Prior, a universal unsupervised denoising framework that reconstructs clean images directly from corrupted observations through a mathematically principled hybrid prior. DNA-Prior integrates (i) an implicit architectural prior, enforced through a deep network parameterisation, with (ii) an explicit spectral-spatial prior composed of a frequency-domain fidelity term and a spatial regularisation functional. This dual-domain formulation yields a well-structured optimisation problem that jointly preserves global frequency characteristics and local anatomical structure, without requiring any external training data or modality-specific tuning. Experiments across multiple modalities show that DNA achieves consistent noise suppression and structural preservation under diverse noise conditions. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_23124 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DNA-Prior: Unsupervised Denoise Anything via Dual-Domain Prior Cheng, Yanqi Cheng, Chun-Wun Denholm, Jim Lima, Thiago Montoya-Zegarra, Javier A. Goodwin, Richard Schönlieb, Carola-Bibiane Aviles-Rivero, Angelica I Computer Vision and Pattern Recognition Medical imaging pipelines critically rely on robust denoising to stabilise downstream tasks such as segmentation and reconstruction. However, many existing denoisers depend on large annotated datasets or supervised learning, which restricts their usability in clinical environments with heterogeneous modalities and limited ground-truth data. To address this limitation, we introduce DNA-Prior, a universal unsupervised denoising framework that reconstructs clean images directly from corrupted observations through a mathematically principled hybrid prior. DNA-Prior integrates (i) an implicit architectural prior, enforced through a deep network parameterisation, with (ii) an explicit spectral-spatial prior composed of a frequency-domain fidelity term and a spatial regularisation functional. This dual-domain formulation yields a well-structured optimisation problem that jointly preserves global frequency characteristics and local anatomical structure, without requiring any external training data or modality-specific tuning. Experiments across multiple modalities show that DNA achieves consistent noise suppression and structural preservation under diverse noise conditions. |
| title | DNA-Prior: Unsupervised Denoise Anything via Dual-Domain Prior |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2511.23124 |