Dictionary learning from incomplete data for efficient image restoration

Naumova, Valeriya; Schnass, Karin

doi:10.23919/eusipco.2017.8081444

Cited by 9 publications

(4 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Yet, its computation time is important and the authors in [15] preferred a simplified but accelerated version called fast EPLL (FEPLL) to reconstruct SEM images [28]. In addition to the patch-based methods used in the microscopy community, wKSVD [29] and ITKrMM [30,31] learn the dictionary from incomplete data without assuming particular patch distribution. They remain state-of-the-art methods that will be considered in this paper.…”

Section: Learning-free Methodsmentioning

confidence: 99%

Fast reconstruction of atomic-scale STEM-EELS images from sparse sampling

et al. 2020

View full text Add to dashboard Cite

This paper discusses the reconstruction of partially sampled spectrum-images to accelerate the acquisition in scanning transmission electron microscopy (STEM). The problem of image reconstruction has been widely considered in the literature for many imaging modalities, but only a few attempts handled 3D data such as spectral images acquired by STEM electron energy loss spectroscopy (EELS). Besides, among the methods proposed in the microscopy literature, some are fast but inaccurate while others provide accurate reconstruction but at the price of a high computation burden. Thus none of the proposed reconstruction methods fulfills our expectations in terms of accuracy and computation complexity. In this paper, we propose a fast and accurate reconstruction method suited for atomic-scale EELS. This method is compared to popular solutions such as beta process factor analysis (BPFA) which is used for the first time on STEM-EELS images. Experiments based on real as synthetic data will be conducted.

show abstract

Section: Learning-free Methodsmentioning

confidence: 99%

Fast reconstruction of atomic-scale STEM-EELS images from sparse sampling

et al. 2020

View full text Add to dashboard Cite

show abstract

“…A sparsity-based sequential method is presented in Algorithm 3 (sequential approach), which consists on learning first the optimal dictionary D and sparse coefficients s i compatible with the incomplete observations (dictionary learning and coding phase), followed by the training phase, where the classifier weights are tuned in order to minimize the classification error of the reconstructed input data vectorsx i = Ds i . It is noted that for the imputation stage (lines 2-12) other and more specialized dictionary learning algorithms with missing data can be applied, such as the ones proposed in [42] for high-dimensional data or [43] for color image data.…”

Section: Supplemental Materialsmentioning

confidence: 99%

Learning from Incomplete Features by Simultaneous Training of Neural Networks and Sparse Coding

Caiafa¹,

Wang

Solé-Casals

et al. 2021

2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

View full text Add to dashboard Cite

In this paper, the problem of training a classifier on a dataset with incomplete features is addressed. We assume that different subsets of features (random or structured) are available at each data instance. This situation typically occurs in the applications when not all the features are collected for every data sample. A new supervised learning method is developed to train a general classifier, such as a logistic regression or a deep neural network, using only a subset of features per sample, while assuming sparse representations of data vectors on an unknown dictionary. Sufficient conditions are identified, such that, if it is possible to train a classifier on incomplete observations so that their reconstructions are well separated by a hyperplane, then the same classifier also correctly separates the original (unobserved) data samples. Extensive simulation results on synthetic and well-known datasets are presented that validate our theoretical findings and demonstrate the effectiveness of the proposed method compared to traditional data imputation approaches and one state-of-the-art algorithm.

show abstract

“…Unavailability and difficulty to generate ground-truth data in various contexts, is the major motivator for unsupervised methods. Noise2Noise [35] and its extensions Noise2Self [5] and Noise2Void [25] demonstrated how denoising can be achieved in an unsupervised manner without clean data. However the aforementioned approaches rely on certain distributional assumptions (i.e.…”

Section: Related Workmentioning

confidence: 99%

“…The main challenge with the data-driven approaches is that finding ground truth for supervision is a hard, time-consuming, usually expensive process and sometimes impossible. Although more recent unsupervised data-driven approaches [35] try to address the ground truth drawback, they rely on assumptions for the noise nature and properties, which do not apply to consumer level depth sensors.…”

Section: Introductionmentioning

confidence: 99%

Self-Supervised Deep Depth Denoising

Sterzentsenko

Saroglou

Chatzitofis

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

View full text Add to dashboard Cite

Depth perception is considered an invaluable source of information for various vision tasks. However, depth maps acquired using consumer-level sensors still suffer from non-negligible noise. This fact has recently motivated researchers to exploit traditional filters, as well as the deep learning paradigm, in order to suppress the aforementioned non-uniform noise, while preserving geometric details. Despite the effort, deep depth denoising is still an open challenge mainly due to the lack of clean data that could be used as ground truth. In this paper, we propose a fully convolutional deep autoencoder that learns to denoise depth maps, surpassing the lack of ground truth data. Specifically, the proposed autoencoder exploits multiple views of the same scene from different points of view in order to learn to suppress noise in a self-supervised endto-end manner using depth and color information during training, yet only depth during inference. To enforce selfsupervision, we leverage a differentiable rendering technique to exploit photometric supervision, which is further regularized using geometric and surface priors. As the proposed approach relies on raw data acquisition, a large RGB-D corpus is collected using Intel RealSense sensors. Complementary to a quantitative evaluation, we demonstrate the effectiveness of the proposed self-supervised denoising approach on established 3D reconstruction applications. Code is avalable at https://github.com/ VCL3D/DeepDepthDenoising

show abstract

Dictionary learning from incomplete data for efficient image restoration

Cited by 9 publications

References 8 publications

Fast reconstruction of atomic-scale STEM-EELS images from sparse sampling

Fast reconstruction of atomic-scale STEM-EELS images from sparse sampling

Learning from Incomplete Features by Simultaneous Training of Neural Networks and Sparse Coding

Self-Supervised Deep Depth Denoising

Contact Info

Product

Resources

About