DEEP-URL: A Model-Aware Approach to Blind Deconvolution Based on Deep Unfolded Richardson-Lucy Network

The lack of interpretability in current deep learning models causes serious concerns as they are extensively used for various life-critical applications. Hence, it is of paramount importance to develop interpretable deep learning models. In this paper, we consider the problem of blind deconvolution and propose a novel model-aware deep architecture that allows for the recovery of both the blur kernel and the sharp image from the blurred image. In particular, we propose the Deep Unfolded Richardson-Lucy (Deep-URL) framework – an interpretable deep-learning architecture that can be seen as an amalgamation of classical estimation technique and deep neural network, and consequently leads to improved performance. Our numerical investigations demonstrate significant improvement compared to state-of-the-art algorithms.

[1]  Qinghua Hu,et al.  Neural Blind Deconvolution Using Deep Priors , 2019, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[2]  Mojtaba Soltanalian,et al.  Model-Aware Deep Architectures for One-Bit Compressive Variational Autoencoding , 2019, ArXiv.

[3]  Ming-Hsuan Yang,et al.  Motion Blur Kernel Estimation via Deep Learning , 2018, IEEE Transactions on Image Processing.

[4]  Ayan Chakrabarti,et al.  A Neural Approach to Blind Motion Deblurring , 2016, ECCV.

[5]  Narendra Ahuja,et al.  A Comparative Study for Single Image Blind Deblurring , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[6]  D. A. Fish,et al.  Blind deconvolution by means of the Richardson-Lucy algorithm. , 1995 .

[7]  Yonina C. Eldar,et al.  An Algorithm Unrolling Approach to Deep Image Deblurring , 2019, ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[8]  Eero P. Simoncelli,et al.  Image quality assessment: from error visibility to structural similarity , 2004, IEEE Transactions on Image Processing.

[9]  Arindam Bose,et al.  Deep Radar Waveform Design for Efficient Automotive Radar Sensing , 2019, ArXiv.

[10]  Frédo Durand,et al.  Understanding and evaluating blind deconvolution algorithms , 2009, 2009 IEEE Conference on Computer Vision and Pattern Recognition.

[11]  Jean-Christophe Pesquet,et al.  Deep unfolding of a proximal interior point method for image restoration , 2018, Inverse Problems.

[12]  Yoshua Bengio,et al.  Gradient-based learning applied to document recognition , 1998, Proc. IEEE.

[13]  Li Xu,et al.  Unnatural L0 Sparse Representation for Natural Image Deblurring , 2013, 2013 IEEE Conference on Computer Vision and Pattern Recognition.

[14]  Daniele Perrone,et al.  A Clearer Picture of Total Variation Blind Deconvolution , 2016, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[15]  Tae Hyun Kim,et al.  Deep Multi-scale Convolutional Neural Network for Dynamic Scene Deblurring , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[16]  Sundaresh Ram,et al.  Removing Camera Shake from a Single Photograph , 2009 .

[17]  Yann LeCun,et al.  Learning Fast Approximations of Sparse Coding , 2010, ICML.

[18]  Deqing Sun,et al.  Deblurring Images via Dark Channel Prior , 2018, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[19]  Jonathan Le Roux,et al.  Deep Unfolding: Model-Based Inspiration of Novel Deep Architectures , 2014, ArXiv.

[20]  Yonina C. Eldar,et al.  Deep Signal Recovery with One-bit Quantization , 2018, ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[21]  Chirag Agarwal,et al.  Removing input features via a generative model to explain their attributions to classifier's decisions , 2019, ArXiv.

[22]  Rob Fergus,et al.  Blind deconvolution using a normalized sparsity measure , 2011, CVPR 2011.

[23]  Li Xu,et al.  Two-Phase Kernel Estimation for Robust Motion Deblurring , 2010, ECCV.

[24]  Anat Levin,et al.  Blind Motion Deblurring Using Image Statistics , 2006, NIPS.