Spatially-Attentive Patch-Hierarchical Network with Adaptive Sampling for Motion Deblurring
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866909099337187328 |
|---|---|
| author | Suin, Maitreya Purohit, Kuldeep Rajagopalan, A. N. |
| author_facet | Suin, Maitreya Purohit, Kuldeep Rajagopalan, A. N. |
| contents | This paper tackles the problem of motion deblurring of dynamic scenes. Although end-to-end fully convolutional designs have recently advanced the state-of-the-art in non-uniform motion deblurring, their performance-complexity trade-off is still sub-optimal. Most existing approaches achieve a large receptive field by increasing the number of generic convolution layers and kernel size. In this work, we propose a pixel adaptive and feature attentive design for handling large blur variations across different spatial locations and process each test image adaptively. We design a content-aware global-local filtering module that significantly improves performance by considering not only global dependencies but also by dynamically exploiting neighboring pixel information. We further introduce a pixel-adaptive non-uniform sampling strategy that implicitly discovers the difficult-to-restore regions present in the image and, in turn, performs fine-grained refinement in a progressive manner. Extensive qualitative and quantitative comparisons with prior art on deblurring benchmarks demonstrate that our approach performs favorably against the state-of-the-art deblurring algorithms. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_06117 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Spatially-Attentive Patch-Hierarchical Network with Adaptive Sampling for Motion Deblurring Suin, Maitreya Purohit, Kuldeep Rajagopalan, A. N. Computer Vision and Pattern Recognition This paper tackles the problem of motion deblurring of dynamic scenes. Although end-to-end fully convolutional designs have recently advanced the state-of-the-art in non-uniform motion deblurring, their performance-complexity trade-off is still sub-optimal. Most existing approaches achieve a large receptive field by increasing the number of generic convolution layers and kernel size. In this work, we propose a pixel adaptive and feature attentive design for handling large blur variations across different spatial locations and process each test image adaptively. We design a content-aware global-local filtering module that significantly improves performance by considering not only global dependencies but also by dynamically exploiting neighboring pixel information. We further introduce a pixel-adaptive non-uniform sampling strategy that implicitly discovers the difficult-to-restore regions present in the image and, in turn, performs fine-grained refinement in a progressive manner. Extensive qualitative and quantitative comparisons with prior art on deblurring benchmarks demonstrate that our approach performs favorably against the state-of-the-art deblurring algorithms. |
| title | Spatially-Attentive Patch-Hierarchical Network with Adaptive Sampling for Motion Deblurring |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2402.06117 |