Stagewise K-SVD to Design Efficient Dictionaries for Sparse Representations

Rusu, Cristian; Dumitrescu, Bogdan

doi:10.1109/lsp.2012.2209871

Cited by 48 publications

(35 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…3 is that the DCT dictionary is well suited for smooth image data and g ∈ {85, 128, 170, 256}. For reference we also show the general dictionary D ∈ R n×n designed by SK-SVD [9]. Sparsity parameters are set to s = 4 and p ∈ {2, 3, 4, 6, 8}.…”

Section: Resultsmentioning

confidence: 99%

“…for which several efficient algorithms that have been proven to perform very well in practice have been proposed [6], [7], [8], [9]. The problem is hard because the objective function is bilinear, i.e., both the dictionary D and the sparse approximation matrix X are unknown, and the constraints are non-convex, i.e., the ℓ 2 equality constraints on the m columns of the dictionary and the ℓ 0 pseudo-norm constraints on the N columns of X.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Learning fast sparsifying overcomplete dictionaries

Rusu

Thompson

2017

2017 25th European Signal Processing Conference (EUSIPCO)

View full text Add to dashboard Cite

Abstract-In this paper we propose a dictionary learning method that builds an overcomplete dictionary that is computationally efficient to manipulate, i.e., sparse approximation algorithms have sub-quadratic computationally complexity. To achieve this we consider two factors (both to be learned from data) in order to design the dictionary: an orthonormal component made up of a fixed number of fast fundamental orthonormal transforms and a sparse component that builds linear combinations of elements from the first, orthonormal component. We show how effective the proposed technique is to encode image data and compare against a previously proposed method from the literature. We expect the current work to contribute to the spread of sparsity and dictionary learning techniques to hardware scenarios where there are hard limits on the computational capabilities and energy consumption of the computer systems.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Learning fast sparsifying overcomplete dictionaries

Rusu

Thompson

2017

2017 25th European Signal Processing Conference (EUSIPCO)

View full text Add to dashboard Cite

show abstract

“…The idea described in this paper is to prune under-utilized, or unused, atoms in the initialization phase and then add new atoms by looking at the worst constructed data items from the dataset using the SVD. The construction of the new atoms is based on ideas from [11]. After the initial dictionary is constructed, the AK-SVD algorithm is applied.…”

Section: Initial Dictionary Constructionmentioning

confidence: 99%

“…The idea of (re)growing a full dictionary from a smaller one is inspired by [11]; however, here the added atoms are not retrained immediately, hence the expansion is very quick.…”

Section: Introductionmentioning

confidence: 99%

An initialization strategy for the dictionary learning problem

Rusu

Dumitrescu

2014

2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

In this paper we present an efficient initialization strategy that improves the performance of overcomplete dictionary learning algorithms. The procedure exploits incoherent structures that can be manipulated and adapted to a given dataset relatively fast. The algorithm involves an iterative adaptation of the dictionary to the dataset with pruning of the less used atoms and constructions of new atoms that fit the data better. Experimental simulations show that the proposed method improves the performance of classical and new developments in dictionary learning algorithms.Index Terms-sparse representations, dictionary learning, initialization.

show abstract

“…The process can be conducted by the following two steps: redundant dictionary design and sparse coefficients solving. Redundant dictionary can be designed by K-SVD algorithm (Rusu and Dumitrescu 2012), shiftinvariant sparse coding algorithm (Plumbley et al 2006), or redundant signal transform basis like wavelet packet basis (Yang et al 2005). Sparse coefficients can be calculated by greedy pursuit algorithms (Bahmani et al 2013), l p norm regularization algorithms (Marjanovic and Solo 2012) and iterative shrinkage algorithms (Beygi et al 2012).…”

Section: Introductionmentioning

confidence: 99%

Intelligent fault diagnosis of rolling element bearings using sparse wavelet energy based on overcomplete DWT and basis pursuit

2015

View full text Add to dashboard Cite

This paper explores sparse time-frequency distribution (TFD) using overcomplete discrete wavelet transform (DWT) and sparse representation techniques. This distribution is discovered for characterizing the periodic transient information embedded in rolling element bearings and extracting effective features that can discriminate different fault conditions. Based on the sparse TFD, a new sparse wavelet energy (SWE) feature is obtained by three main steps: first, an overcomplete discrete DWT is employed to decompose the fault signal and construct a redundant dictionary; second, the redundant dictionary is optimized by basis pursuit to obtain the sparsest TFD; finally, SWE is calculated from the new TFD to produce a feature vector for each signal. SWE features that combine the merits of overcomplete DWT and sparse representation techniques can precisely reveal fault-induced information, thereby exhibiting valuable properties for automatic fault identification by intelligent classifiers. The effectiveness and advantages of the proposed features are confirmed by simulation and the practical fault pattern recognition of rolling bearings.

show abstract

Stagewise K-SVD to Design Efficient Dictionaries for Sparse Representations

Cited by 48 publications

References 11 publications

Learning fast sparsifying overcomplete dictionaries

Learning fast sparsifying overcomplete dictionaries

An initialization strategy for the dictionary learning problem

Intelligent fault diagnosis of rolling element bearings using sparse wavelet energy based on overcomplete DWT and basis pursuit

Contact Info

Product

Resources

About