Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2507.17892 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912499096354816 |
|---|---|
| author | Liu, Hanzhou Li, Binghan Liu, Chengkai Lu, Mi |
| author_facet | Liu, Hanzhou Li, Binghan Liu, Chengkai Lu, Mi |
| contents | Transformers, with their self-attention mechanisms for modeling long-range dependencies, have become a dominant paradigm in image restoration tasks. However, the high computational cost of self-attention limits scalability to high-resolution images, making efficiency-quality trade-offs a key research focus. To address this, Restormer employs channel-wise self-attention, which computes attention across channels instead of spatial dimensions. While effective, this approach may overlook localized artifacts that are crucial for high-quality image restoration. To bridge this gap, we explore Dilated Neighborhood Attention (DiNA) as a promising alternative, inspired by its success in high-level vision tasks. DiNA balances global context and local precision by integrating sliding-window attention with mixed dilation factors, effectively expanding the receptive field without excessive overhead. However, our preliminary experiments indicate that directly applying this global-local design to the classic deblurring task hinders accurate visual restoration, primarily due to the constrained global context understanding within local attention. To address this, we introduce a channel-aware module that complements local attention, effectively integrating global context without sacrificing pixel-level precision. The proposed DiNAT-IR, a Transformer-based architecture specifically designed for image restoration, achieves competitive results across multiple benchmarks, offering a high-quality solution for diverse low-level computer vision problems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_17892 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DiNAT-IR: Exploring Dilated Neighborhood Attention for High-Quality Image Restoration Liu, Hanzhou Li, Binghan Liu, Chengkai Lu, Mi Computer Vision and Pattern Recognition Transformers, with their self-attention mechanisms for modeling long-range dependencies, have become a dominant paradigm in image restoration tasks. However, the high computational cost of self-attention limits scalability to high-resolution images, making efficiency-quality trade-offs a key research focus. To address this, Restormer employs channel-wise self-attention, which computes attention across channels instead of spatial dimensions. While effective, this approach may overlook localized artifacts that are crucial for high-quality image restoration. To bridge this gap, we explore Dilated Neighborhood Attention (DiNA) as a promising alternative, inspired by its success in high-level vision tasks. DiNA balances global context and local precision by integrating sliding-window attention with mixed dilation factors, effectively expanding the receptive field without excessive overhead. However, our preliminary experiments indicate that directly applying this global-local design to the classic deblurring task hinders accurate visual restoration, primarily due to the constrained global context understanding within local attention. To address this, we introduce a channel-aware module that complements local attention, effectively integrating global context without sacrificing pixel-level precision. The proposed DiNAT-IR, a Transformer-based architecture specifically designed for image restoration, achieves competitive results across multiple benchmarks, offering a high-quality solution for diverse low-level computer vision problems. |
| title | DiNAT-IR: Exploring Dilated Neighborhood Attention for High-Quality Image Restoration |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2507.17892 |