Deep Image Prior

@article{Ulyanov2018DeepIP,
  title={Deep Image Prior},
  author={Dmitry Ulyanov and Andrea Vedaldi and Victor S. Lempitsky},
  journal={2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year={2018},
  pages={9446-9454}
}
Deep convolutional networks have become a popular tool for image generation and restoration. [...] Key Method Furthermore, the same prior can be used to invert deep neural representations to diagnose them, and to restore images based on flash-no flash input pairs. Apart from its diverse applications, our approach highlights the inductive bias captured by standard generator network architectures. It also bridges the gap between two very popular families of image restoration methods: learning-based methods using…Expand
Deep Image Prior
TLDR
It is shown that a randomly-initialized neural network can be used as a handcrafted prior with excellent results in standard inverse problems such as denoising, super-resolution, and inpainting. Expand
Image Denoising Using a Novel Deep Generative Network with Multiple Target Images and Adaptive Termination Condition
TLDR
This work proposes a novel deep generative network with multiple target images and an adaptive termination condition, enabling better guidance during the convergence process and improving the convergence speed, and adopted the noise level estimation (NLE) technique to set a more reasonable adaptive terminate condition, which can effectively solve the problem of over-fitting. Expand
Image Restoration by Deep Projected GSURE
TLDR
This paper proposes a new image restoration framework that is based on minimizing a loss function that includes a ”projected-version” of the Generalized Stein Unbiased Risk Estimator (GSURE) and parameterization of the latent image by a CNN. Expand
Deep Decoder: Concise Image Representations from Untrained Non-convolutional Networks
TLDR
This paper proposes an untrained simple image model, called the deep decoder, which is a deep neural network that can generate natural images from very few weight parameters, with a simple architecture with no convolutions and fewer weight parameters than the output dimensionality. Expand
Blind Image Restoration with Flow Based Priors
TLDR
This work proposes using normalizing flows to model the distribution of the target content and to use this as a prior in a maximum a posteriori (MAP) formulation and presents experimental results for a number of different degradations on data sets varying in complexity. Expand
NAS-DIP: Learning Deep Image Prior with Neural Architecture Search
TLDR
This work proposes to search for neural architectures that capture stronger image priors by leveraging an existing neural architecture search algorithm (using reinforcement learning with a recurrent neural network controller) and designs new search spaces for an upsampling cell and a pattern of cross-scale residual connections. Expand
Generic Image Restoration with Flow Based Priors
TLDR
This work proposes using normalizing flows to model the distribution of the target content and to use this as a prior in a maximum a posteriori (MAP) formulation and presents experimental results for a number of different degradations on data sets varying in complexity. Expand
Synthetic Images as a Regularity Prior for Image Restoration Neural Networks
TLDR
This paper shows that deep neural networks can reach near-optimal performances by training them on a synthetic dataset made of realizations of a dead leaves model, both for image denoising and super-resolution. Expand
Image Processing Using Multi-Code GAN Prior
TLDR
A novel approach is proposed, called mGANprior, to incorporate the well-trained GANs as effective prior to a variety of image processing tasks, by employing multiple latent codes to generate multiple feature maps at some intermediate layer of the generator and composing them with adaptive channel importance to recover the input image. Expand
Rate-optimal denoising with deep neural networks
TLDR
This paper considers the problem of denoising an image from additive Gaussian noise using the two generator based approaches, and state and analyze a simple gradient algorithm that minimizes a non-convex loss function, and provably reduces noise energy by a factor of O(k/n). Expand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 34 REFERENCES
Non-local Color Image Denoising with Convolutional Neural Networks
  • Stamatios Lefkimmiatis
  • Computer Science
  • 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2017
TLDR
This work proposes a novel deep network architecture for grayscale and color image denoising that is based on a non-local image model and highlights a direct link of the proposed non- local models to convolutional neural networks. Expand
Shepard Convolutional Neural Networks
TLDR
This paper draws on Shepard interpolation and design Shepard Convolutional Neural Networks (ShCNN) which efficiently realizes end-to-end trainable TVI operators in the network and shows that by adding only a few feature maps in the new Shepard layers, the network is able to achieve stronger results than a much deeper architecture. Expand
Visualizing Deep Convolutional Neural Networks Using Natural Pre-images
TLDR
This paper studies several landmark representations, both shallow and deep, by a number of complementary visualization techniques based on the concept of “natural pre-image”, and shows that several layers in CNNs retain photographically accurate information about the image, with different degrees of geometric and photometric invariance. Expand
EnhanceNet: Single Image Super-Resolution Through Automated Texture Synthesis
TLDR
This work proposes a novel application of automated texture synthesis in combination with a perceptual loss focusing on creating realistic textures rather than optimizing for a pixelaccurate reproduction of ground truth images during training to achieve a significant boost in image quality at high magnification ratios. Expand
Image Super-Resolution via Deep Recursive Residual Network
TLDR
This paper proposes a very deep CNN model (up to 52 convolutional layers) named Deep Recursive Residual Network (DRRN) that strives for deep yet concise networks, and recursive learning is used to control the model parameters while increasing the depth. Expand
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network
  • C. Ledig, Lucas Theis, +6 authors W. Shi
  • Computer Science, Mathematics
  • 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2017
TLDR
SRGAN, a generative adversarial network (GAN) for image super-resolution (SR), is presented, to its knowledge, the first framework capable of inferring photo-realistic natural images for 4x upscaling factors and a perceptual loss function which consists of an adversarial loss and a content loss. Expand
Learning a Deep Convolutional Network for Image Super-Resolution
TLDR
This work proposes a deep learning method for single image super-resolution (SR) that directly learns an end-to-end mapping between the low/high-resolution images and shows that traditional sparse-coding-based SR methods can also be viewed as a deep convolutional network. Expand
Deep Laplacian Pyramid Networks for Fast and Accurate Super-Resolution
TLDR
This paper proposes the Laplacian Pyramid Super-Resolution Network (LapSRN) to progressively reconstruct the sub-band residuals of high-resolution images and generates multi-scale predictions in one feed-forward pass through the progressive reconstruction, thereby facilitates resource-aware applications. Expand
Understanding deep image representations by inverting them
Image representations, from SIFT and Bag of Visual Words to Convolutional Neural Networks (CNNs), are a crucial component of almost any image understanding system. Nevertheless, our understanding ofExpand
Globally and locally consistent image completion
We present a novel approach for image completion that results in images that are both locally and globally consistent. With a fully-convolutional neural network, we can complete images of arbitraryExpand
...
1
2
3
4
...