Supervised Learning of Image Restoration with Convolutional Networks

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

doi:10.1109/iccv.2007.4408909

Cited by 183 publications

(161 citation statements)

References 19 publications

Supporting

Mentioning

161

Contrasting

Order By: Relevance

“…This problem can, however, be solved (Helmstaedter et al, unpublished) by combining high-accuracy long-range manual annotation, as reported here, with locally accurate but globally error-prone automated volume reconstructions [24][25][26] .…”

Section: Combining Skeletons With Automated Methodsmentioning

confidence: 99%

See 1 more Smart Citation

High-accuracy neurite reconstruction for high-throughput neuroanatomy

2011

Self Cite

View full text Add to dashboard Cite

Neuroanatomic analysis depends on the reconstruction of complete cell shapes. Highthroughput reconstruction of neural circuits ("connectomics") using volume electron microscopy requires dense staining of all cells, where even experts make annotation errors.Currently, reconstruction rather than acquisition speed limits the determination of neural wiring diagrams. We present methods for the fast and reliable reconstruction of densely labeled datasets. Our approach, based on manually skeletonizing each neurite redundantly (multiple times) with a special visualization/annotation software tool (KNOSSOS), is ~50 times faster than volume labeling. Errors are detected and eliminated by a "redundantskeleton consensus procedure" (RESCOP), which uses a statistical model of how true neurite connectivity is transformed into annotation decisions. RESCOP also estimates the consensus skeletons' reliability. Focused re-annotation of difficult locations promises a rather steep increase of reliability as a function of the average skeleton redundancy and thus the nearly error-free analysis of large neuroanatomical datasets.

show abstract

Section: Combining Skeletons With Automated Methodsmentioning

confidence: 99%

“…Computer algorithms, especially those using machine learning [24][25][26] can help with the reconstruction of neural circuits. In the long run, such tools may well replace or at least greatly reduce the need for manual annotation.…”

Section: Evaluation Of Automated Reconstruction Algorithmsmentioning

confidence: 99%

High-accuracy neurite reconstruction for high-throughput neuroanatomy

2011

Self Cite

View full text Add to dashboard Cite

show abstract

“…For this purpose, cell boundary enhancement (and eventual segmentation) has attracted attention in recent years [6,18]. Focus of the work in this field has been on largely unsupervised diffusion based techniques [17,12] as well as those based on bulky convolutional networks [5,6] and graph-cuts methods [18]. In order to emphasize the usefulness of RadonLike features, we would restrict ourselves to unsupervised methods.…”

Section: Cell Boundary Enhancementmentioning

confidence: 99%

Radon-Like features and their application to connectomics

Kumar

Vázquez-Reina

Pfister

2010

2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Workshops

View full text Add to dashboard Cite

In this paper we present a novel class of so-called Radon-Like features, which allow for aggregation of spatially distributed image statistics into compact feature descriptors. Radon-Like features, which can be efficiently computed, lend themselves for use with both supervised and unsupervised learning methods. Here we describe various instantiations of these features and demonstrate there usefulness in context of neural connectivity analysis, i.e. Connectomics, in electron micrographs. Through various experiments on simulated as well as real data we establish the efficacy of the proposed features in various tasks like cell membrane enhancement, mitochondria segmentation, cell background segmentation, and vesicle cluster detection as compared to various other state-of-the-art techniques.

show abstract

“…In related work, Jain et al uses supervised learning to classify pixels as membrane or nonmembrane in specimens prepared with an extracellular stain [3]. This stain shows only the cell boundaries and results in simpler structures.…”

Section: Introductionmentioning

confidence: 99%

An optimal-path approach for neural circuit reconstruction

Jurrus

Whitaker

Jones

et al. 2008

2008 5th IEEE International Symposium on Biomedical Imaging: From Nano to Macro

View full text Add to dashboard Cite

Neurobiologists are collecting large amounts of electron microscopy image data to gain a better understanding of neuron organization in the central nervous system. Image analysis plays an important role in extracting the connectivity present in these images; however, due to the large size of these datasets, manual analysis is essentially impractical. Automated analysis, however, is challenging because of the difficulty in reliably segmenting individual neurons in 3D. In this paper, we describe an automatic method for finding neurons in sequences of 2D sections. The proposed method formulates the problem of finding paths through sets of sections as an optimal path computation, which applies a cost function to the identification of a cell from one section to the next and solves this optimization problem using Dijkstra's algorithm. This basic formulation allows us to account for variability or inconsistencies between sections and to prioritize cells based on the evidence of their connectivity.

show abstract

Supervised Learning of Image Restoration with Convolutional Networks

Cited by 183 publications

References 19 publications

High-accuracy neurite reconstruction for high-throughput neuroanatomy

High-accuracy neurite reconstruction for high-throughput neuroanatomy

Radon-Like features and their application to connectomics

An optimal-path approach for neural circuit reconstruction

Contact Info

Product

Resources

About