Fjölbreytumerkjafræði - Multivariate signal processing - verkefni lokið

Fréttatilkynning verkefnisstjóra

26.1.2017

This project marked the beginning of a multivariate signal processing laboratory at the University of Iceland that focuses on the development of signal and image processing algorithms.

This project aimed at developing signal processing methods for analyzing multivariate data.  The focus was on signal processing models that assume that the high dimensional data to be investigated lives in a low dimensional subspace.  The goal was to develop algorithms that characterize this subspace and estimate the relevant parameters.

Although that the applicability of the models and procedures considered in this project is wide-ranging we primarily focused on the following application areas: Hyperspectral unmixing, processing of hyperspectral data, and to lesser extent brain imaging.

Heiti verkefnis: Fjölbreytumerkjafræði / Multivariate signal processing
Verkefnisstjóri: Magnús Örn Úlfarsson, Háskóla Íslands
Tegund styrks: Verkefnisstyrkur
Styrkár: 2013-2015
Fjárhæð styrks: 21,78 millj. kr.
Tilvísunarnúmer Rannís: 130635-05

The results from the project have been reported in 8 high impact ISI journal papers, 2 PhD theses and numerous conference papers. Also, this project marked the beginning of a multivariate signal processing laboratory (www.csi.hi.is)  at the University of Iceland that focuses on the development of signal and image processing algorithms.  This project has also produced a lot of computer code that implements the method developed.  We have distributed the computer code by request basis. 

Publications

Journal Articles

 [1] M.O. Ulfarsson, V. Solo, Tuning Parameter Selection for Underdetermined Reduced-Rank Regression, IEEE Signal Processing Letters, volume 20, issue 9, pages 881-884, 2013.

 [2] J. Sigurdsson, M.O. Ulfarsson, and J.R. Sveinsson, Hyperspectral unmixing using $l_q$ penalty, IEEE Transactions on Geoscience and Remote Sensing, volume 52, issue 11, pages 6793-6806, 2014.

[3] B. Rasti, J.R. Sveinsson, and M.O. Ulfarsson, Wavelet Based Sparse Reduced Rank Regression for Hyperspectral Image Restoration,  IEEE Transactions on Geoscience and Remote Sensing, volume 52, issue 10, pages 6688- 6698, 2014.

[4] M.O. Ulfarsson, V.Solo, Selecting the Number of Principal Components with SURE, IEEE Signal Processing Letters, Volume 22, Issue 2, 239-243,2015.

[5] B. Rasti, M.O. Ulfarsson, and J.R. Sveinsson, Hyperspectral Subspace Identification Using SURE, IEEE Geoscience and Remote Sensing Letters, volume 12, issue 12, pages2481 - 2485, year 2015.

[6] M.O. Ulfarsson, F. Palsson, and J. Sigurdsson, Classification of Big Data With Application to Imaging Genetics, Proceedings of the IEEE,  volume 104, number 11, pages 2137 - 2154, 2016.

[7] Jakob Sigurdsson; Magnus Orn Ulfarsson; Johannes R. Sveinsson, Blind Hyperspectral Unmixing Using Total Variation and  $l_q$ Sparse Regularization, IEEE Transactions on Geoscience and Remote Sensing, volume 54, number 11, pages  6371 - 6384, 2016.

[8] Behnood Rasti; Magnus Orn Ulfarsson; Johannes R. Sveinsson, Hyperspectral Feature Extraction Using Total Variation Component Analysis, IEEE Transactions on Geoscience and Remote Sensing, volume 54, number 12, pages 6976 - 6985, 2016.

Peer reviewed conference articles

[9] M.O. Ulfarsson and V. Solo, ”'Tuning Parameter Selection for Nonnegative Matrix Factorization”, In Proc IEEE. International Conference on Acoustics, Speech and Signal Processing (ICASSP 2013), Vancouver, Canada, 2013.

[10] J. Sigurdsson, M.O. Ulfarsson, and J.R. Sveinsson, Smooth and Sparse Hyperspectral Unmixing using an l0 penalty, IEEE Workshop on Hyperspectral Image and Signal Processing Gainsville, Florida, USA, 2013.

[11] J. Sigurdsson, M. O. Ulfarsson, J. R. Sveinsson, and J. A. Benediktsson, Smooth Spectral Unmixing using Total Variation Regularization and a First Order Roughness Penalty, IEEE International Conference on Geoscience and Remote Sensing, (IGARSS 2013), Melbourne, AUSTRALIA, 2013.

[12] J. Sigurdsson, M.O. Ulfarsson, J.R. Sveinsson, and J.A. Benediktsson, Sparse Representation of Hyperspectral Data using CUR Matrix Decomposition, IEEE International Conference on Geoscienceand Remote Sensing (IGARSS 2013), Melbourne, AUSTRALIA, 2013.

[13] B. Rasti, J.R. Sveinsson, M.O. Ulfarsson and J. Sigurdsson, First Order Roughness Penalty for Hyperspectral Image Denoising, IEEE Workshop on Hyperspectral Image and Signal Processing Gainsville, Florida, USA, 2013.

[14] B. Rasti, J.R. Sveinsson, M.O. Ulfarsson and J. Sigurdsson, Wavelet Based Sparse Principal Component Analysis for Hyperspectral Denoising, IEEE Workshop on Hyperspectral Image and Signal Processing Gainsville, Florida, USA, 2013.

[15] B. Rasti, J.R. Sveinsson, M.O. Ulfarsson, and J.A. Benediktsson, Hyperspectral Image Denoising Using a New Linear Model and Sparse Regularization, IEEE International Conference on Geoscience and Remote Sensing, (IGARSS 2013), Melbourne, AUSTRALIA, 2013.

[16] B. Rasti, J. R. Sveinsson, M. O. Ulfarsson and J. A. Benediktsson, Hyperspectral image restoration using wavelets, SPIE Remote Sensing, Dresden, Germany, 2013.

[17] B. Rasti, J. R. Sveinsson, M. O. Ulfarsson and J. A. Benediktsson, Wavelet based hyperspectral image restoration using spatial and spectral penalties, SPIE Remote Sensing, Dresden, Germany, 2013.

[18] M.O. Ulfarsson and V. Solo, ”Sparse Component Analysis via Dyadic Cyclic Descent”, In Proc IEEE. International Conference on Acoustics, Speech and Signal Processing (ICASSP 2014), Florence, Italy, 2014.

[19] F. Palsson, M.O. Ulfarsson and J.R. Sveinsson, ”Sparse Gaussian Noisy Independent Component Analysis”, In Proc IEEE. International Conference on Acoustics, Speech and Signal Processing (ICASSP 2014), Florence, Italy, 2014.

[20] F. Palsson, M.O. Ulfarsson, and J.R. Sveinsson, "HYPERSPECTRAL IMAGE DENOISING USING A SPARSE LOW RANK MODEL AND DUAL-TREE COMPLEX WAVELET TRANSFORM", IEEE International Conference on Geoscience and Remote Sensing, (IGARSS 2014), Quebec City, Canada, 2014.

[21] B. Rasti, M.O. Ulfarsson and J.R. Sveinsson, "SURE BASED MODEL SELECTION FOR HYPERSPECTRAL IMAGING", IEEE International Conference on Geoscience and Remote Sensing, (IGARSS 2014), Quebec City, Canada, 2014.

[22] B. Rasti, J.R. Sveinsson and M.O. Ulfarsson, "TOTAL VARIATION BASED HYPERSPECTRAL FEATURE EXTRACTION", IEEE International Conference on Geoscience and Remote Sensing, (IGARSS 2014), Quebec City, Canada, 2014.

[23] J. Sigurdsson, M.O. Ulfarsson and J.R. Sveinsson, "Semi-supervised hyperspectral unmixing", IEEE International Conference on Geoscience and Remote Sensing, (IGARSS 2014), Quebec City, Canada, 2014.

[24] J. Sigurdsson, M.O. Ulfarsson and J.R. Sveinsson, "Endmember Constrained Semi-Supervised Hyperspectral Unmixing", IEEE Workshop on Hyperspectral Image and Signal Processing, Lausanne, Switzerland, 2014.

[25] M.O. Ulfarsson, V. Solo, G. Marjanovic, " Sparse and Low Rank Decomposition using l0 penalty", IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP15), Brisbane, 2015.

 [26] J. Sigurdsson, M.O. Ulfarsson and J.R. Sveinsson, "Total Variation and lq Based Hyperspectral Unmixing for Feature Extraction and Classification", IEEE International Conference on Geoscience and Remote Sensing, (IGARSS 2015), Milano,  2015.

PhD Theses

[27] Behnood Rasti, Sparse Hyperspectral image modeling and restoration, University of Iceland, 2014

[28] Jakob Sigurðsson, Hyperspectral Unmixing Using Total Variation and Sparse Methods, University of Iceland 2015.

 

 

 









Þetta vefsvæði byggir á Eplica