Mask-Specific Inpainting with Deep Neural Networks

By Rolf Köhler, Christian J. Schuler, Bernhard Schölkopf, Stefan Harmeling. 2014

125 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Large Scale Image Completion via Co-Modulated Generative Adversarial Networks (2021, background): “Then, (Köhler et al., 2014; Ren et al., 2015; Xie et al., 2012) adopt deep neural networks for image completion; (Pathak et al., 2016) first exploits conditional GANs.”
Generative Image Inpainting with Contextual Attention (2018, background): “Initial efforts [21, 37] train convolutional neural networks for denoising and inpainting of small regions.”
Pluralistic Image Completion (2019, background): “Initial works [20, 30] focused on small and thin holes.”
PEPSI++: Fast and Lightweight Network for Image Inpainting (2019, methodology): “The CNN-based image inpainting methods employing an encoder–decoder structure have shown superior performance on inpainting the complex hole region compared with the diffusion- or patch-based methods [8], [18].”
Prior Guided GAN Based Semantic Inpainting (2020, methodology): “Single-pass inference models → Initial frameworks of [24, 43] were mainly deep regression networks trained with the usual 2 reconstruction loss.”
Shift-Net: Image Inpainting via Deep Feature Rearrangement (2018, background): “Originally, CNN-based inpainting is confined to small and thin masks [21, 34, 45].”
Globally and locally consistent image completion (2017, methodology): “Initially, CNN-based image inpainting approaches were limited to very small and thin masks [Köhler et al. 2014; Ren et al. 2015; Xie et al. 2012].”
PEPSI : Fast Image Inpainting With Parallel Decoding Network (2019, methodology): “Thanks to a decade of advances in CNNs, image inpainting methods adopting an encoder-decoder structure have achieved a significant progress [13, 24].”
Image Inpainting With Learnable Bidirectional Attention Maps (2019, methodology): “Early CNN-based methods [14, 21, 30] are suggested for handling images with small and thin holes.”
Sparsity Invariant CNNs (2017, background): “[28] pass an additional binary validity mask to the network.”
Sparse Gradient Regularized Deep Retinex Network for Robust Low-Light Image Enhancement (2021, background): “These applications include denoising [25], completion [26], inpainting [27], super-resolution [28], deblurring [29], deconvolution [30], and style transfer [31], etc.”
Learning Proximal Operators: Using Denoising Networks for Regularizing Inverse Imaging Problems (2017, background): “ImageNet classification [24, 35, 17], deep networks have recently been extended to high dimensional inverse problems such as image denoising [43, 47], deblurring [44], superresolution [10, 11], optical flow estimation [12, 28], image demosaicking [42, 15, 21], or inpainting [22, 45].”