Online multimodal dictionary learning

Traoré, Abraham; Bérar, Maxime; Rakotomamonjy, Alain

doi:10.1016/j.neucom.2019.08.053

Cited by 17 publications

(18 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Constraints, which impose smoothness in time and nonnegativity or sparsity, may also be included depending on the application context, such as, for example, in nonnegative online CPD for time-evolving topic modeling (with possible applications in the analysis of social media-generated data on the Covid-19 pandemic) [27] or in tensor dictionary learning [41]. Possible solution approaches include alternating optimization with alternating direction method of multipliers (AO-ADMM) [42].…”

Section: Related Workmentioning

confidence: 99%

“…Applications of OTF abound. They include unveiling the topology of evolving networks [70], spatio-temporal prediction or image in-painting [41], multiple-input multiple-output (MIMO) wireless communications [13], [71], brain imaging [72], monitoring heart-related features from wearable sensors for multi-lead electro-cardiography (ECG) [73], anomaly detection in networks and topic modeling [16], structural health monitoring (in an internet of things (IoT) context) [36], online cartography (spectrum map reconstruction in cognitive radio networks) [14], detection of anomalies in the process of 3D printing [74], data traffic monitoring in networks [10], [16], cardiac MRI [10], stream data compression (e.g., in power distribution systems [75] or in video [76]), and online completion [10], [77], [78], among others.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Online Rank-Revealing Block-Term Tensor Decomposition

Rontogiannis¹,

Kofidis²,

Giampouras³

2021

Preprint

View full text Add to dashboard Cite

The so-called block-term decomposition (BTD) tensor model, especially in its rank-(Lr, Lr, 1) version, has been recently receiving increasing attention due to its enhanced ability of representing systems and signals that are composed of block components of rank higher than one, a scenario encountered in numerous and diverse applications. Its uniqueness and approximation have thus been thoroughly studied. The challenging problem of estimating the BTD model structure, namely the number of block terms (rank) and their individual (block) ranks, is of crucial importance in practice and has only recently started to attract significant attention. In data-streaming scenarios and/or big data applications, where the tensor dimension in one of its modes grows in time or can only be processed incrementally, it is essential to be able to perform model selection and computation in a recursive (incremental/online) manner. To date there is only one such work in the literature concerning the (general rank-(L, M, N )) BTD model, which proposes an incremental method, however with the BTD rank and block ranks assumed to be a-priori known and time invariant. In this preprint, a novel approach to rank-(Lr, Lr, 1) BTD model selection and tracking is proposed, based on the idea of imposing column sparsity jointly on the factors and estimating the ranks as the numbers of factor columns of nonnegligible magnitude. An online method of the alternating iteratively reweighted least squares (IRLS) type is developed and shown to be computationally efficient and fast converging, also allowing the model ranks to change in time. Its time and memory efficiency are evaluated and favorably compared with those of the batch approach. Simulation results are reported that demonstrate the effectiveness of the proposed scheme in both selecting and tracking the correct BTD model.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Online Rank-Revealing Block-Term Tensor Decomposition

Rontogiannis¹,

Kofidis²,

Giampouras³

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Grussler et al verified that the rank-constrained Frobenius norm can outperform the nuclear norm [23]. Traoré et al proposed to enforce a sparsity constraint on the core tensor and low-rank ( F  norm) constraints on factor matrices, and learnt reconstructed patch for inpainting images using the factor matrices and the core tensor [24]. The approach was also flexible to incorporate nonnegativity and orthogonality constraints [24].…”

Section: Introductionmentioning

confidence: 99%

“…Traoré et al proposed to enforce a sparsity constraint on the core tensor and low-rank ( F  norm) constraints on factor matrices, and learnt reconstructed patch for inpainting images using the factor matrices and the core tensor [24]. The approach was also flexible to incorporate nonnegativity and orthogonality constraints [24]. Bahri et al proposed Kronecker component analysis (RKCA) method for image denoising.…”

Section: Introductionmentioning

confidence: 99%

Low-Rank Tucker-2 Model for Multi-Subject fMRI Data Decomposition With Spatial Sparsity Constraint

Yue

Lin

Kuang

et al. 2022

IEEE Trans. Med. Imaging

View full text Add to dashboard Cite

Tucker decomposition can provide an intuitive summary to understand brain function by decomposing multisubject fMRI data into a core tensor and multiple factor matrices, and was mostly used to extract functional connectivity patterns across time/subjects using orthogonality constraints. However, these algorithms are unsuitable for extracting common spatial and temporal patterns across subjects due to distinct characteristics such as high-level noise. Motivated by a successful application of Tucker decomposition to image denoising and the intrinsic sparsity of spatial activations in fMRI, we propose a low-rank Tucker-2 model with spatial sparsity constraint to analyze multisubject fMRI data. More precisely, we propose to impose a sparsity constraint on spatial maps by using an p  norm (0

show abstract

“…In [15] and [16], authors compute the CPD of successive sub-tensors in an adaptive way but they assume that the rank is known and does not vary with time while in [17] the sub-tensors are treated independently using AutoTen for rank estimation. Eventually, in [18] authors introduced an Online Tucker decomposition.…”

Section: Introductionmentioning

confidence: 99%

Online Nonnegative Canonical Polyadic Decomposition: Algorithms and Application

Sanou

Redon

Luciani

et al. 2021

2021 29th European Signal Processing Conference (EUSIPCO)

View full text Add to dashboard Cite

The Nonnegative Canonical Polyadic Decomposition (NN-CPD) is now widely used in signal processing to decompose multi-way arrays thanks to non-negative factor matrices. In many applications, a three way array is built from collections of 2D-signals and new signals are regularly recorded. In this case one may want to update the factor matrices after each new measurement without computing the NN-CPD of the whole array. We then speak of Online NN-CPD. In this context the main difficulty is that the number of relevant factors is unknown and can vary with time. In this paper we propose two algorithms to compute the Online NN-CPD based on sparse dictionary learning. We then introduce an application example of Online NN-CPD in environmental sciences and evaluate the performances of the proposed approach in this context on real data.

show abstract

Online multimodal dictionary learning

Cited by 17 publications

References 22 publications

Online Rank-Revealing Block-Term Tensor Decomposition

Online Rank-Revealing Block-Term Tensor Decomposition

Low-Rank Tucker-2 Model for Multi-Subject fMRI Data Decomposition With Spatial Sparsity Constraint

Online Nonnegative Canonical Polyadic Decomposition: Algorithms and Application

Contact Info

Product

Resources

About