analiza aiqbench
This research introduces a novel dual-domain architecture that combines Vision Transformers with an FFT-ReLU module to address computational bottlenecks in high-resolution image deblurring. By explicitly bridging spatial attention modeling with frequency sparsity, the model effectively suppresses blur artifacts while preserving fine details outperforming current state-of-the-art benchmarks on standard datasets. The work represents a significant methodological advance for computer vision systems requiring efficient and accurate restoration of motion-blurred imagery.
From Attention to Frequency: Integration of Vision Transformer and FFT-ReLU for Enhanced Image Deblurring
arXiv:2511.10806v1 Announce Type: cross Abstract: Image deblurring is vital in computer vision, aiming to recover sharp images from blurry ones caused by motion or camera shake. While deep learning approaches such as CNNs and Vision Transformers (ViTs) have advanced this field, they often struggle with complex or high-resolution blur and computational demands. We propose a new dual-domain architecture that unifies Vision Transformers with a frequency-domain FFT-ReLU module, explicitly bridging spatial attention modeling and frequency sparsity. In this structure, the ViT backbone captures local and global dependencies, while the FFT-ReLU component enforces frequency-domain sparsity to suppress blur-related artifacts and preserve fine details. Extensive experiments on benchmark datasets demonstrate that this architecture achieves superior PSNR, SSIM, and perceptual quality compared to state-of-the-art models. Both quantitative metrics, qualitative comparisons, and human preference evaluations confirm its effectiveness, establishing a practical and generalizable paradigm for real-world image restoration.
Abstract:Image deblurring is vital in computer vision, aiming to recover sharp images from blurry ones caused by motion or camera shake. While deep learning approaches such as CNNs and Vision Transformers (ViTs) have advanced this field, they often struggle with complex or high-resolution blur and computational demands. We propose a new dual-domain architecture that unifies Vision Transformers with a frequency-domain FFT-ReLU module, explicitly bridging spatial attention modeling and frequency sparsity. In this structure, the ViT backbone captures local and global dependencies, while the FFT-ReLU component enforces frequency-domain sparsity to suppress blur-related artifacts and preserve fine details. Extensive experiments on benchmark datasets demonstrate that this architecture achieves superior PSNR, SSIM, and perceptual quality compared to state-of-the-art models. Both quantitative metrics, qualitative comparisons, and human preference evaluations confirm its effectiveness, establishing a practical and generalizable paradigm for real-world image restoration.
Image and Video Processing (eess.IV); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
arXiv:2511.10806 [eess.IV]
(or arXiv:2511.10806v1 [eess.IV] for this version)
Proceedings of the 18th International Conference on Agents and Artificial Intelligence (ICAART 2026), Volume 2, Marbella, Spain, March 5-7, 2026, pp. 1810-1820. SCITEPRESS
https://doi.org/10.5220/0014441500004052 DOI(s) linking to related resources
Text integral sub licența deschisă a publicației · 192 cuvinte
Subiecte conexe
EA-LiteUNet: An Edge-Adaptive and Resource-Efficient U-Net for Boundary-Sensitive Dermoscopic Image Segmentation
arXiv · 18 aug. 2026
BRICS Draws a New Global Line: Full Text of New Delhi Declaration
slguardian.org · 12 sept. 2026
[AINews] The Future of Latent Space
Latent Space · 25 sept. 2026
Humanity’s Last Invention — Richard Socher of Recursive
Latent Space · 14 sept. 2026