Zou, Zhentao ; Chen, Lin ; Jiang, Xue ; Zoubir, Abdelhak. M. (2024)
Exploiting Generative Diffusion Prior With Latent Low-Rank Regularization for Image Inpainting.
In: IEEE Signal Processing Letters, 31
doi: 10.1109/LSP.2024.3453665
Artikel, Bibliographie
Kurzbeschreibung (Abstract)
Generative diffusion models have recently shown impressive results in image restoration. However, the predicted noise from existing diffusion-based methods may be inaccurate, especially when the noise amplitude is small, thereby leading to sub-optimal results. In this letter, an unsupervised diffusion model with latent low-rank regularization is proposed to alleviate this challenge. In particular, we first create a latent low-rank space using self-supervised learning for each degraded images, from which we derive corresponding latent low-rank regularization. This regularization, combining with observed prior information and smoothness regularization, guides the reserve sampling process, resulting in the generation of high-quality images with fine-grained textures and fewer artifacts. In addition, by utilizing the pre-trained unconditional diffusion model, the proposed model reconstructs the missing pixels in a zero-shot manner, which does not need any reference images for additional training. Extensive experimental results demonstrate that our proposed method is superior to the self-supervised tensor completion methods and representative diffusion model-based image restoration methods.
| Typ des Eintrags: | Artikel |
|---|---|
| Erschienen: | 2024 |
| Autor(en): | Zou, Zhentao ; Chen, Lin ; Jiang, Xue ; Zoubir, Abdelhak. M. |
| Art des Eintrags: | Bibliographie |
| Titel: | Exploiting Generative Diffusion Prior With Latent Low-Rank Regularization for Image Inpainting |
| Sprache: | Englisch |
| Publikationsjahr: | 3 September 2024 |
| Verlag: | IEEE |
| Titel der Zeitschrift, Zeitung oder Schriftenreihe: | IEEE Signal Processing Letters |
| Jahrgang/Volume einer Zeitschrift: | 31 |
| DOI: | 10.1109/LSP.2024.3453665 |
| Kurzbeschreibung (Abstract): | Generative diffusion models have recently shown impressive results in image restoration. However, the predicted noise from existing diffusion-based methods may be inaccurate, especially when the noise amplitude is small, thereby leading to sub-optimal results. In this letter, an unsupervised diffusion model with latent low-rank regularization is proposed to alleviate this challenge. In particular, we first create a latent low-rank space using self-supervised learning for each degraded images, from which we derive corresponding latent low-rank regularization. This regularization, combining with observed prior information and smoothness regularization, guides the reserve sampling process, resulting in the generation of high-quality images with fine-grained textures and fewer artifacts. In addition, by utilizing the pre-trained unconditional diffusion model, the proposed model reconstructs the missing pixels in a zero-shot manner, which does not need any reference images for additional training. Extensive experimental results demonstrate that our proposed method is superior to the self-supervised tensor completion methods and representative diffusion model-based image restoration methods. |
| Fachbereich(e)/-gebiet(e): | 18 Fachbereich Elektrotechnik und Informationstechnik 18 Fachbereich Elektrotechnik und Informationstechnik > Institut für Nachrichtentechnik 18 Fachbereich Elektrotechnik und Informationstechnik > Institut für Nachrichtentechnik > Signalverarbeitung |
| Hinterlegungsdatum: | 27 Sep 2024 09:39 |
| Letzte Änderung: | 27 Sep 2024 09:39 |
| PPN: | |
| Export: | |
| Suche nach Titel in: | TUfind oder in Google |
 |
Frage zum Eintrag |
Optionen (nur für Redakteure)
 |
Redaktionelle Details anzeigen |
Drucken
Impressum