Supervised Learning of Image Restoration with Convolutional Networks

By Viren Jain, Joseph F. Murray, Fabian Roth, Srinivas C. Turaga, Valentin P. Zhigulin, Kevin L. Briggman, Moritz Helmstaedter, Winfried Denk, H. Sebastian Seung. 2007

253 citations recorded by the source corpus.

Subfields: Artificial Intelligence, 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 Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Deep multi-scale video prediction beyond mean square error (2015, background): “The simplest and oldest one is to have no pooling/subsampling but many convolution layers (Jain et al., 2007).”
Conditional Random Fields as Recurrent Neural Networks (2015, positive): “opredictstructuredoutput are found in different domains. For example, Do et al. [14] proposed an approach to combine deep neural networks and Markov networks for sequence labeling tasks. Jain et al . [26] has shown Convolutional Neural Networks can perform well like MRFs/CRFs approaches in image res”
N4-Fields: Neural Network Nearest Neighbor Fields for Image Transforms (2014, background): “Finally, a large body of work, e.g. [13, 3] simply frame the segmentation problem as patch classiﬁcation, making generic CNN-based classiﬁcation easily applicable and successful.”
Deep learning for computational biology (2016, positive): “(Li, 2009) in such raw data analysis tasks, for example restoring noisy neural circuitry images (Jain et al , 2007).”
Predicting Citywide Crowd Flows Using Deep Spatio-Temporal Residual Networks (2017, methodology): “Being different from the classical CNN, we do not use subsampling, but only convolutions [9].”
Projection onto the Manifold of Elongated Structures for Accurate Extraction (2015, methodology): “For example, a Convolutional Neural Network and a hierarchical segmentation framework combining Random Forest classifier and watersheds are used in [19] and [3] respectively to segment neural membranes.”
3D Convolutional Neural Networks for Human Action Recognition (2010, unknown): “In addition to recognition tasks, CNNs have also been used in 3D image restoration problems [32].”
Long-Term Image Boundary Extrapolation (2016, background): “The output layer neurons should have a wide receptive field [8] to preserve long range spatial and temporal dependencies and learn about interaction with other boundaries in a spatiotemporal context.”
Deep Convolutional Neural Networks for Multi-Modality Isointense Infant Brain Image Segmentation (2015, methodology): “In Jain et al. (2007), Jain and Seung (2009), Turaga et al. (2010), and Helmstaedter et al. (2013), CNNs were applied to restore and segment the volumetric electron microscopy images.”
Training energy-based models for time-series imputation (2013, methodology): “Convolutional neural networks have been trained to restore images (Jain et al., 2007) in a supervised way but in that work the task was not to impute missing values but to undo the effects of a contaminating process in a way that is more similar to denoising.”
Semantic Image Segmentation with Contextual Hierarchical Models (2015, methodology): “Jain et al. [42], Turaga et al. [43], and Ciresan et al. [44] used convnets for membrane detection and cell segmentation in EM images.”
Deep Spatio-Temporal Residual Networks for Citywide Crowd Flows Prediction (2016, methodology): “Being different from the classical CNN, we do not use subsampling, but only convolutions (Jain et al. 2007).”
Unsupervised Deep Learning Applied to Breast Density Segmentation and Mammographic Risk Scoring (2016, unknown): “They were found to be faster and more expressive than other graphical models like Markov or Conditional Random Fields [47].”