Online convolutional dictionary learning

Liu, Jialin; Garcia-Cardona, Cristina; Wohlberg, Brendt; Yin, Wotao

doi:10.1109/icip.2017.8296573

Cited by 23 publications

(37 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another aspect of this advantage is our ability to run in an online manner, even for a single input image. This stands in sharp contrast to other recent online methods [30,31] which allow for online training but only in the case of streaming images. Other approaches took a step further and proposed partitioning the image into smaller sub-images [32], but this is still far from [32] and the SBDL algorithm [1].…”

Section: Relation To Other Methodsmentioning

confidence: 66%

A Local Block Coordinate Descent Algorithm for the CSC Model

Zisselman

Sulam²,

Elad

2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

View full text Add to dashboard Cite

The Convolutional Sparse Coding (CSC) model has recently gained considerable traction in the signal and image processing communities. By providing a global, yet tractable, model that operates on the whole image, the CSC was shown to overcome several limitations of the patch-based sparse model while achieving superior performance in various applications. Contemporary methods for pursuit and learning the CSC dictionary often rely on the Alternating Direction Method of Multipliers (ADMM) in the Fourier domain for the computational convenience of convolutions, while ignoring the local characterizations of the image. A recent work by Papyan et al. [1] suggested the SBDL algorithm for the CSC, while operating locally on image patches. SBDL demonstrates better performance compared to the Fourier-based methods, albeit still relying on the ADMM.In this work we maintain the localized strategy of the SBDL, while proposing a new and much simpler approach based on the Block Coordinate Descent algorithm -this method is termed Local Block Coordinate Descent (LoBCoD). Furthermore, we introduce a novel stochastic gradient descent version of LoBCoD for training the convolutional filters. The Stochastic-LoBCoD leverages the benefits of online learning, while being applicable to a single training image. We demonstrate the advantages of the proposed algorithms for image inpainting and multi-focus image fusion, achieving state-of-the-art results.

show abstract

Section: Relation To Other Methodsmentioning

confidence: 66%

A Local Block Coordinate Descent Algorithm for the CSC Model

Zisselman

Sulam²,

Elad

2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

View full text Add to dashboard Cite

show abstract

“…The Fast Iterative Shrinkage-Thresholding Algorithm (FISTA) [33], an accelerated proximal gradient method, has been used for CSC [6], [5], [19], and in a recent online CDL algorithm [18], but has not previously been considered for the dictionary update of a batch-mode dictionary learning algorithm.…”

Section: D / Frequency Domain Consensusmentioning

confidence: 99%

“…All the results using the methods discussed and analyzed in the main document were computed using the Python implementation of the SPORCO library [36], [ also include comparisons with the method proposed by Papyan et al [27], using their publicly available Matlab and C implementation 16 . We tried to include the publicly available Matlab implementations of the methods proposed by Šorel and Šroubek 17 [24] and by Heide et al 18 [9] in these comparisons, but were unable 16 to obtain acceptable results 19 . We therefore omit these methods from the comparisons here, including them only in a separate set of experiments on a smaller data set, reported in Sec.…”

Section: Sv Large Training Set Experimentsmentioning

confidence: 99%

Convolutional Dictionary Learning: A Comparative Review and New Algorithms

Garcia-Cardona

Wohlberg

2018

IEEE Trans. Comput. Imaging

Self Cite

157

114

View full text Add to dashboard Cite

Convolutional sparse representations are a form of sparse representation with a dictionary that has a structure that is equivalent to convolution with a set of linear filters. While effective algorithms have recently been developed for the convolutional sparse coding problem, the corresponding dictionary learning problem is substantially more challenging. Furthermore, although a number of different approaches have been proposed, the absence of thorough comparisons between them makes it difficult to determine which of them represents the current state of the art. The present work both addresses this deficiency and proposes some new approaches that outperform existing ones in certain contexts. A thorough set of performance comparisons indicates a very wide range of performance differences among the existing and proposed methods, and clearly identifies those that are the most effective. We do not consider the analysis form [2] of sparse representations in this work, focusing instead on the more common synthesis form. 2 We do not consider the very recent online CDL algorithms [18], [19], [20], [21] in this work.

show abstract

“…Thus, Theorem 1 also holds for Algorithm 3, and a stationary point of problem (15) can be obtained. E. Discussion with [27], [28] Here, we discuss the very recent works of [27], [28] which also consider online learning of the dictionary in CSC. Extra experiments are performed in Section B, which shows our method is much faster than them.…”

Section: Convergencementioning

confidence: 99%

“…However, though their O(·) are the same, storing sparse matrix requires 2-3 times more space than storing dense matrix with the same number of nonzero entries, as each nonzero entry in a sparse matrix needs to be kept in the compressed sparse row (CSR) 3 format [30]. Moreover, though [28] and the proposed method take O(N K 2 P ) time to update the history matrices, using sparse matrices as in [28] is empirically slower [30]. Preliminary experiments show that with K = 100 filters, the proposed algorithm is 5 times faster on an P = 100 × 100 image and 10 times faster on an P = 200 × 200 image.…”

Section: Convergencementioning

confidence: 99%

Scalable Online Convolutional Sparse Coding

Wang

Yao

Kwok

et al. 2018

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

Convolutional sparse coding (CSC) improves sparse coding by learning a shift-invariant dictionary from the data. However, most existing CSC algorithms operate in the batch mode and are computationally expensive. In this paper, we alleviate this problem by online learning. The key is a reformulation of the CSC objective so that convolution can be handled easily in the frequency domain, and much smaller history matrices are needed. To solve the resultant optimization problem, we use the alternating direction method of multipliers (ADMMs), and its subproblems have efficient closed-form solutions. Theoretical analysis shows that the learned dictionary converges to a stationary point of the optimization problem. Extensive experiments are performed on both the standard CSC benchmark data sets and much larger data sets such as the ImageNet. Results show that the proposed algorithm outperforms the state-of-the-art batch and online CSC methods. It is more scalable, has faster convergence, and better reconstruction performance.

show abstract

Online convolutional dictionary learning

Cited by 23 publications

References 14 publications

A Local Block Coordinate Descent Algorithm for the CSC Model

A Local Block Coordinate Descent Algorithm for the CSC Model

Convolutional Dictionary Learning: A Comparative Review and New Algorithms

Scalable Online Convolutional Sparse Coding

Contact Info

Product

Resources

About