Learning Deep Gradient Descent Optimization for Image Deconvolution

  title={Learning Deep Gradient Descent Optimization for Image Deconvolution},
  author={Dong Gong and Zhen Zhang and Qinfeng Shi and Anton van den Hengel and Chunhua Shen and Yanning Zhang},
  journal={IEEE Transactions on Neural Networks and Learning Systems},
As an integral component of blind image deblurring, non-blind deconvolution removes image blur with a given blur kernel, which is essential but difficult due to the ill-posed nature of the inverse problem. [...] Key Method We propose a Recurrent Gradient Descent Network (RGDN) by systematically incorporating deep neural networks into a fully parameterized gradient descent scheme. A parameter-free update unit is used to generate updates from the current estimates, based on a convolutional neural network.Expand
A Deep Optimization Approach for Image Deconvolution
The Golf Optimizer is proposed, a novel but simple form of network that learns deep priors from data with better propagation behavior, and first estimates an aggressive propagation towards optimum using one network and recurrently applies a residual CNN to learn the gradient of prior for delicate correction on restoration. Expand
Deep Wiener Deconvolution: Wiener Meets Deep Learning for Image Deblurring
The proposed deep Wiener deconvolution network facilitates deblurred results with visibly fewer artifacts, and quantitatively outperforms state-of-the-art non-blind image deblurring methods by a wide margin. Expand
Learning Spatially-Variant MAP Models for Non-blind Image Deblurring
The classical maximum a-posteriori (MAP) framework for non-blind image deblurring requires defining suitable data and regularization terms, whose interplay yields the desired clear image throughExpand
MPTV: Matching Pursuit-Based Total Variation Minimization for Image Deconvolution
The inhomogeneity of MPTV alleviates the over-smoothing and ringing artifacts and improves the robustness to errors in blur kernel, and is less sensitive to the choice of the trade-off parameter between data fitting and regularization. Expand
Attention-Guided Network for Ghost-Free High Dynamic Range Imaging
The proposed AHDRNet is a non-flow-based method, which can also avoid the artifacts generated by optical-flow estimation error, and can achieve state-of-the-art quantitative and qualitative results. Expand
Deep Outlier Handling for Image Deblurring
This work proposes to learn a deep convolutional neural network to directly estimate the confidence map, which can identify reliable inliers and outliers from the blurred image and thus facilitates the following deblurring process. Expand
Learning to Zoom-In via Learning to Zoom-Out: Real-World Super-Resolution by Generating and Adapting Degradation
A framework to learn SR from an arbitrary set of unpaired LR and HR images and to see how far a step can go in such a realistic and “unsupervised” setting is proposed. Expand
Regularization-based modulation transfer function compensation for optical satellite image restoration using joint statistical model in curvelet domain
This work exploits the merit of image prior characteristics in both local smoothness and nonlocal self-similarity properties of an image in a hybrid domain to design effective regularization terms that reflect these image properties. Expand
Generalized Nesterov Accelerated Conjugate Gradient Algorithm for a Compressively Sampled MR Imaging Reconstruction
The proposed GNACG_TV algorithm can better reconstruct under-sampled MR images than other 4 methods and can further improve the convergence speed based on Nesterov acceleration and get better reconstruction performance. Expand
Transductive Relation-Propagation With Decoupling Training for Few-Shot Learning.
  • Yuqing Ma, Shihao Bai, +5 authors Meng Wang
  • Medicine
  • IEEE transactions on neural networks and learning systems
  • 2021
A transductive relation-propagation graph neural network (GNN) with a decoupling training strategy (TRPN-D) to explicitly model and propagate such relations across support-query pairs, and empower the few-shot module the ability of transferring past knowledge to new tasks via the decoupled training. Expand


Deep Convolutional Neural Network for Image Deconvolution
This work develops a deep convolutional neural network to capture the characteristics of degradation, establishing the connection between traditional optimization-based schemes and a neural network architecture where a novel, separable structure is introduced as a reliable support for robust deconvolution against artifacts. Expand
Blind Image Deconvolution by Automatic Gradient Activation
A gradient activation method is introduced to automatically select a subset of gradients of the latent image in a cutting-plane-based optimization scheme for kernel estimation, which greatly improves the accuracy and flexibility and affords great convenience for handling noise and outliers. Expand
Learning Fully Convolutional Networks for Iterative Non-blind Deconvolution
The proposed fully convolutional network is able to learn an adaptive image prior, which keeps both local and global information, and performs favorably against state-of-the-art algorithms in terms of quality and speed. Expand
A Machine Learning Approach for Non-blind Image Deconvolution
This work relies on a two-step procedure, but learns the second step on a large dataset of natural images, using a neural network, and shows that this approach outperforms the current state-of-the-art on aLarge dataset of artificially blurred images. Expand
From Motion Blur to Motion Flow: A Deep Learning Solution for Removing Heterogeneous Motion Blur
  • Dong Gong, Jie Yang, +5 authors Qinfeng Shi
  • Computer Science, Mathematics
  • 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2017
This work directly estimates the motion flow from the blurred image through a fully-convolutional deep neural network (FCN) and recovers the unblurred image from the estimated motion flow and is the first universal end-to-end mapping from the blur image to the dense motion flow. Expand
Learning Deep CNN Denoiser Prior for Image Restoration
Experimental results demonstrate that the learned set of denoisers can not only achieve promising Gaussian denoising results but also can be used as prior to deliver good performance for various low-level vision applications. Expand
Good Image Priors for Non-blind Deconvolution - Generic vs. Specific
Two strategies that become beneficial as example images become more context-appropriate are discovered: locally adapted priors trained from region level correspondence significantly outperform globally trained priors, and a novel multi-scale patch-pyramid formulation is more successful at transferring mid and high frequency details from example scenes. Expand
One Network to Solve Them All — Solving Linear Inverse Problems Using Deep Projection Models
This work proposes a general framework to train a single deep neural network that solves arbitrary linear inverse problems and demonstrates superior performance over traditional methods using wavelet sparsity prior while achieving performance comparable to specially-trained networks on tasks including compressive sensing and pixel-wise inpainting. Expand
A Generic Deep Architecture for Single Image Reflection Removal and Image Smoothing
A deep neural network structure that exploits edge information in addressing representative low-level vision tasks such as layer separation and image filtering by estimating edges and reconstructing images using only cascaded convolutional layers arranged such that no handcrafted or application-specific image-processing components are required. Expand
Understanding and evaluating blind deconvolution algorithms
The previously reported failure of the naive MAP approach is explained by demonstrating that it mostly favors no-blur explanations and it is shown that since the kernel size is often smaller than the image size a MAP estimation of the kernel alone can be well constrained and accurately recover the true blur. Expand