NFResNet: Multi-scale and U-shaped Networks for Deblurring
Abstract
Multi-Scale and U-shaped Networks are widely used in various image restoration problems, including deblurring. Keeping in mind the wide range of applications, we present a comparison of these architectures and their effects on image deblurring. We also introduce a new block called as NFResblock. It consists of a Fast Fourier Transformation layer and a series of modified Non-Linear Activation Free Blocks. Based on these architectures and additions, we introduce NFResnet and NFResnet+, which are modified multi-scale and U-Net architectures, respectively. We also use three different loss functions to train these architectures: Charbonnier Loss, Edge Loss, and Frequency Reconstruction Loss. Extensive experiments on the Deep Video Deblurring dataset, along with ablation studies for each component, have been presented in this paper. The proposed architectures achieve a considerable increase in Peak Signal to Noise (PSNR) ratio and Structural Similarity Index (SSIM) value.
Cite
@article{arxiv.2212.05909,
title = {NFResNet: Multi-scale and U-shaped Networks for Deblurring},
author = {Tanish Mittal and Preyansh Agrawal and Esha Pahwa and Aarya Makwana},
journal= {arXiv preprint arXiv:2212.05909},
year = {2023}
}
Comments
Due to limitations in GPU Compute, We weren't able to test the paper on the popularly used GoPro Dataset which is mostly used for testing image deblurring problems. Afterwards the submission on Arxiv, We observed that we missed comparison of our results with some State-of-the-art papers like ARVo & Gated Spatio-Temporal Attention-Guided Video Deblurring