Deblur image using transformers that modulate self-attention outputs dynamically with DeblurDiNAT
Deblur image using transformers that modulate self-attention outputs dynamically with DeblurDiNAT
DeblurDiNAT: A Lightweight and Effective Transformer for Image Deblurring
arXiv paper abstract https://arxiv.org/abs/2403.13163
arXiv PDF paper https://arxiv.org/pdf/2403.13163.pdf
Blurry images may contain local and global non-uniform artifacts, which complicate the deblurring process and make it more challenging to achieve satisfactory results.
... propose DeblurDiNAT, a compact encoder-decoder Transformer which efficiently restores clean images from real-world blurry ones.
... adopt an alternating dilation factor structure with the aim of global-local feature learning.
... propose a channel modulation self-attention (CMSA) block, where a cross-channel learner (CCL) is utilized to capture channel relationships.
... present a divide and multiply feed-forward network (DMFN) allowing fast feature propagation ... design a lightweight gated feature fusion (LGFF) module, which performs controlled feature merging.
... DeblurDiNAT, provides ... achieves state-of-the-art (SOTA) ... on ... image deblurring datasets ... space-efficient and time-saving method ... stronger generalization ... 3%-68% fewer parameters ... deblurred ... closer to the ground truth.
Please like and share this post if you enjoyed it using the buttons at the bottom!
Stay up to date. Subscribe to my posts https://morrislee1234.wixsite.com/website/contact
Web site with my other posts by category https://morrislee1234.wixsite.com/website
Comments