Fields of expertise
computational harmonic analysis
LinksLTS2 lab home page
MissionData nowadays come in overwhelming volume. In order to cope with this deluge, we explore and use the benefits of geometry and symmetry in higher dimensional data. But volume is not the only problem: data models are also increasingly complex, mixing various components. We thus use redundant dictionaries as a dimensionality reduction tool to dig out information from complicated high-dimensional datasets and multichannel signals, or to model complex behaviours in more classical signals. Finally data can also be complex because they are collected on surfaces, or more generally manifolds, or because they are not scalar-valued. We thus explore extensions of Computational Harmonic Analysis in higher dimensions, in complex geometries or for non-scalar data.
Visit our lab web pages.
BiographyPierre Vandergheynst received the M.S. degree in physics and the Ph.D. degree in mathematical
physics from the Université catholique de Louvain, Louvain-la-Neuve, Belgium, in 1995 and 1998, respectively. From 1998 to 2001, he was a Postdoctoral Researcher with the Signal Processing Laboratory, Swiss Federal Institute of Technology (EPFL), Lausanne, Switzerland. He was Assistant Professor at EPFL (2002-2007), where he is now a Full Professor of Electrical Engineering and, by courtesy, of Computer and Communication Sciences. As of 2015, Prof. Vandergheynst serves as EPFL’s Vice-Provost for Education.
His research focuses on harmonic analysis, sparse approximations and mathematical data processing in general with applications covering signal, image and high dimensional data processing, computer vision, machine learning, data science and graph-based data processing.
He was co-Editor-in-Chief of Signal Processing (2002-2006), Associate Editor of the IEEE Transactions on Signal Processing (2007-2011), the flagship journal of the signal processing community and currently serves as Associate Editor of Computer Vision and Image Understanding and SIAM Imaging Sciences. He has been on the Technical Committee of various conferences, serves on the steering committee of the SPARS workshop and was co-General Chairman of the EUSIPCO 2008 conference.
Pierre Vandergheynst is the author or co-author of more than 70 journal papers, one monograph and several book chapters. He has received two IEEE best paper awards. Professor Vandergheynst is a laureate of the Apple 2007 ARTS award and of the 2009-2010 De Boelpaepe prize of the Royal Academy of Sciences of Belgium.
Recent publications (2009-present)
Generalised Implicit Neural Representations2022. 36th Conference on Neural Information Processing Systems (NeurIPS 2022), New Orleans, Louisiana, USA, November 28 - December 9, 2022.
Social Network Architectures of Disinformation2020-08-14
Musical Source Separation2020-08-14
On the Experimental Transferability of Spectral Graph Convolutional Networks2020-06-19
Geometric deep learning for medium-range weather prediction2020-06-19
Fourier could be a data scientist: From graph Fourier transform to signal processing on graphsComptes Rendus Physique. 2019-07-01. DOI : 10.1016/j.crhy.2019.08.003.
Spherical Convolutionnal Neural Networks: Empirical Analysis of SCNNs2019-06-21
Learning Representations of Source Code from Structure and Context2019-03-15
Tensor Robust Pca On Graphs2019-01-01. 44th IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Brighton, ENGLAND, May 12-17, 2019. p. 5406-5410. DOI : 10.1109/ICASSP.2019.8682990.
Graph Laplacians for Rotation Equivariant Convolutional Neural Networks2019
Spectrally approximating large graphs with smaller graphs2018-07-10. International Conference in Machine Learning (ICML), Stockholmsmässan, Sweden, July 10-15, 2018.
Computational Thinking and Thinking ComputationallyDia-logos, Ramon Llull's Method of Thought and Artistic Practice; Minneapolis: University of Minnesota Press, 2018.
Joint Estimation Of The Room Geometry And Modes With Compressed Sensing2018-01-01. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Calgary, CANADA, Apr 15-20, 2018. p. 6882-6886. DOI : 10.1109/ICASSP.2018.8462655.
Adaptive Graph-Based Total Variation for Tomographic ReconstructionsIEEE SIGNAL PROCESSING LETTERS. 2018. DOI : 10.1109/LSP.2018.2816582.
Graph Signal Processing: Overview, Challenges and ApplicationsProceedings of the IEEE. 2018. DOI : 10.1109/JPROC.2018.2820126.
Fast Approximate Spectral Clustering for Dynamic Networks2018. International Conference in Machine Learning (ICML), Stockholmsmässan, Sweden, July 10-15, 2018.
Random sampling of bandlimited signals on graphsApplied and Computational Harmonic Analysis. 2018. DOI : 10.1016/j.acha.2016.05.005.
Towards Stationary Time-Vertex Signal Processing2017. IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), New Orleans, LA, MAR 05-09, 2017. p. 3914-3918. DOI : 10.1109/ICASSP.2017.7952890.
System, device, and method for contextual knowledge retrieval and displayWO2017187401 . 2017.
Transient networks of spatio-temporal connectivity map communication pathways in brain functional systemsNeuroimage. 2017. DOI : 10.1016/j.neuroimage.2017.04.015.
Compressive Embedding and Visualization using Graphs2017
Structured Sequence Modeling with Graph Convolutional Recurrent Networks2017.
FMA: A Dataset For Music Analysis2017. 18th International Society for Music Information Retrieval Conference, Suzhou, China, October 23-28, 2017.
An Algorithm Architecture Co-Design for CMOS Compressive High Dynamic Range ImagingIeee Transactions On Computational Imaging. 2016. DOI : 10.1109/Tci.2016.2557073.
Multilinear Low-Rank Tensors on Graphs & Applications2016.
Compressed Sensing and Adaptive Graph Total Variation for Tomographic ReconstructionsIEEE Transactions on Medical Imaging. 2016.
Tensor low-rank and sparse light field photographyComputer Vision And Image Understanding. 2016. DOI : 10.1016/j.cviu.2015.11.004.
Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering2016. Advances in Neural Information Processing Systems 29, Barcelona, Spain, December 5-10, 2016.
Compressive Spectral Clustering2016. 33rd International Conference on Machine Learning (ICML), New York, USA, June 19-24. p. 1002-1011.
Source Localization on Graphs via l1 Recovery and Spectral Graph Theory2016. 12th IEEE Image, Video, and Multidimensional Signal Processing (IVMSP) Workshop 2016, Bordeaux, France, July 11-12, 2016. DOI : 10.1109/IVMSPW.2016.7528230.
Low-Rank Matrices on Graphs: Generalized Recovery & Applications2016
Localisation et identification spectrale conjointe de sources large bande par parcimonie groupée2016. Congrès Français d'Acoustique, Le Mans, France, April 2016.
Graph Based Sinogram Denoising for Tomographic Reconstructions2016.
Learning Laplacian Matrix in Smooth Graph Signal RepresentationsIEEE Transactions on Signal Processing. 2016. DOI : 10.1109/TSP.2016.2602809.
Dynamic activation patterns in brain MRI data2015. 25th Colloque Gretsi, Lyon.
Mapping resting-state dynamics on spatio-temporal graphs: a combined functional and diffusion MRI approach2015. 23rd International Symposium on Magnetic Resonance in Medicine (ISMRM), Toronto.
Structured Auto-Encoder with application to Music Genre Recognition2015
Random Sampling of Bandlimited Signals on GraphsNIPS2015 Workshop on Multiresolution Methods for Large Scale Learning, Montréal, December 12th, 2015.
Learning class-specific descriptors for deformable shapes using localized spectral convolutional networksComputer Graphics Forum. 2015. DOI : 10.1111/cgf.12693.
Random sampling of bandlimited signals on graphs2015
A Convex Solution to Disparity Estimation from Light Fields via the Primal-Dual Method2015. 10th International Conference on Energy Minimization Methods in Computer Vision and Pattern Recognition (EMMCVPR), Hong Kong, PEOPLES R CHINA, JAN 13-16, 2015. p. 350-363. DOI : 10.1007/978-3-319-14612-6_26.
Accelerated filtering on graphs using Lanczos methodIEEE Signal Processing Letters. 2015.
ShapeNet: Convolutional Neural Networks on Non-Euclidean Manifolds2015.
Enhanced Matrix Completion with Manifold LearningInternational BASP Frontiers Workshop 2015, Villars-sur-Ollon, Switzerland, January 25 - 30, 2015.
Spectrum-Adapted Tight Graph Wavelet and Vertex-Frequency FramesIEEE Transactions on Signal Processing. 2015. DOI : 10.1109/Tsp.2015.2424203.
Graph-based Image Inpainting2014
System and method for media library navigation and recommendationWO2014002064 . 2014.
Audio Steganography using Convex Demixing2014
Matrix Completion on GraphsUCL - Duke Workshop on Sensing and Analysis of High-Dimensional Data (SAHD 2014), London, England, September 4-5, 2014.
Matrix Completion on Graphs2014. Neural Information Processing Systems 2014, Workshop "Out of the Box: Robustness in High Dimension", Montreal, Canada, December 8-13, 2014.
A Convex Solution to Disparity Estimation from Light Fields via the Primal-Dual Method2014. Energy Minimization Methods in Computer Vision and Pattern Recognition (EMMCVPR), Hong Kong, China.
Robust visual tracking using feature selection2014
Numerical experiments with MALDI Imaging dataAdvances In Computational Mathematics. 2014. DOI : 10.1007/s10444-013-9325-0.
Approximate Compressed Sensing: Ultra-Low Power Biosignal Processing via Aggressive Voltage Scaling on a Hybrid Memory Multi-core Processor2014. International Symposium on Low Power Electronics and Design (ISLPED 2014), La Jolla, California, USA, August 11-13, 2014. p. 40-45. DOI : 10.1145/2627369.2627629.
Ultra low power design of hardware efficient CS-Based compression scheme in WBSN2014. CT-Energy Community Workshop, Barcelona, Spain, April 23-25, 2014.
Hardware-Software Inexactness in Noise-aware Design of Low-Power Body Sensor Nodes2014. Designing with Uncertainty - Opportunities & Challenges, York, United Kingdom, March 17-19 , 2014.
A Compressed Sensing Framework for Magnetic Resonance FingerprintingSiam Journal On Imaging Sciences. 2014. DOI : 10.1137/130947246.
Compressed Quantitative MRI: Bloch Response Recovery through iterated projection2014. IEEE International Conference on Accoustics, Speech and Signal Processing (ICASSP), Florence, Italy. DOI : 10.1109/ICASSP.2014.6854937.
Compact Low-Power Cortical Recording Architecture for Compressive Multichannel Data AcquisitionIEEE Transactions on Biomedical Circuits and Systems. 2014. DOI : 10.1109/TBCAS.2014.2304582.
Robust Real-time Pedestrians Detection in Urban Environments with a Network of Low Resolution CamerasTransportation Research Part C: Emerging Technologies. 2014. DOI : 10.1016/j.trc.2013.11.019.
Clustering on Multi-Layer Graphs via Subspace Analysis on Grassmann ManifoldsIEEE Transactions on Signal Processing. 2014. DOI : 10.1109/TSP.2013.2295553.
Method to compensate the effect of the rolling shutter effectUS8350922 ; US2011267514 . 2013.
Joint image registration and reconstruction from compressed multi-view measurements2013. Conference on Wavelets and Sparsity XV. DOI : 10.1117/12.2023916.
On the Sparsity of Wavelet Coefficients for Signals on Graphs2013. Conference on Wavelets and Sparsity XV. DOI : 10.1117/12.2022850.
The Emerging Field of Signal Processing on Graphs: Extending High-Dimensional Data Analysis to Networks and Other Irregular DomainsIEEE Signal Processing Magazine. 2013. DOI : 10.1109/Msp.2012.2235192.
Traitement du signal sur les graphes2013 GRETSI Symposium, Brest, France, September 3-6, 2013.
compressive source separation for hyperspectral imaging2013
High Frame-Rate Low-Power Compressive Sampling CMOS Image Sensor Architecture23rd Great Lakes Symposium on VLSI, Paris, France, May 2-3, 2013.
Power-Efficient CMOS Image Acquisition System based on Compressive Sampling2013. 56th IEEE International Midwest Symposium on Circuits and Systems, Columbus, Ohio, USA, August 4-7, 2013. p. 1367-1370. DOI : 10.1109/MWSCAS.2013.6674910.
Column-Separated Compressive Sampling Scheme for Low Power CMOS Image Sensors2013. 11th IEEE International New Circuits and Systems (NEWCAS) Conference, Paris, France, June 16-19, 2013. DOI : 10.1109/NEWCAS.2013.6573620.
Sparse Binary Features for Image Classification2013
Joint reconstruction of misaligned images from incomplete measurements for cardiac MRI2013. 10th International Conference on Sampling Theory and Applications (SAMPTA), Bremen, Germany, July 2013.
Non-convex optimization for robust multi-view imagingInternational Biomedical and Astronomical Signal Processing (BASP) Frontiers workshop, Villars-sur-Ollon, Switzerland, January, 2013.
Multichannel Blind Deconvolution Using Low-rank and Sparse Decomposition2013. SPARS, 2013.
Image reconstruction of non-planar scenes from compressed multi-view measurements2013. Workshop on Signal Processing with Adaptive Sparse Structured Representations (SPARS 2013), Lausanne, 2013.
Non-convex optimization for robust multi-view imaging2013. International Biomedical and Astronomical Signal Processing (BASP) Frontiers workshop, Villars-sur-Ollon, Switzerland, January, 2013.
Omnidirectional sensor array systemUS10362225 ; US2018220070 ; US9876953 ; US2014146132 ; WO2012056437 . 2012.
Automatic online delineation of a multi-lead electrocardiogram bio signalUS2014148714 ; EP2654557 ; WO2012085841 . 2012.
Image transform for video codingUS8300693 ; US2006159179 . 2012.
Analyse de données en grande dimension sur graphes et réseauxJournée Mathématiques et Grandes Dimensions, Polytech Lyon, December 10, 2012.
Compressed Sensing of Simultaneous Low-Rank and Joint-Sparse MatricesIEEE Transactions on Information Theory. 2012.
Compressed Sensing of Simultaneous Low-Rank and Joint-Sparse Matrices2012
Robust joint reconstruction of misaligned images using semi-parametric dictionariesICML Workshop on Sparsity, Dictionaries and Projections in Machine Learning and Signal Processing, Edinburgh, June 30, 2012.
Robust joint reconstruction of misaligned images using semi-parametric dictionariesWorkshop on Sparsity, Localization and Dictionary Learning, London, June 26, 2012.
On localisation and uncertainty measures on graphs2012
Beyond Bits: Reconstructing Images from Local Binary Descriptors2012. 21st International Conference on Pattern Recognition (ICPR), Tsukuba Science City, Japan, November 11-15, 2012. p. 935-938.
Robust joint reconstruction of misaligned images using semi-parametric dictionaries2012. ICML Workshop on Sparsity, Dictionaries and Projections in Machine Learning and Signal Processing, Edinburgh, Scotland, June 30, 2012.
A Real-time Multi-camera System with Omnidirectional Image Reconstruction CapabilityICCP2012, Seattle, WA, USA, April 27-29, 2012.
Light Field Compressive Sensing2012. 1st International Traveling Workshop for Interacting Sparse Models and Technology, Marseille, France.
Light Field Tensor Recovery2012. IEEE International Conference on Image Processing, Orlando, Florida, USA.
Audio-Visual Object Extraction using Graph CutsIEEE Transactions on Image Processing. 2012.
Sparse Approximation Using M-Term Pursuit and Application in Image and Video CodingIEEE Transactions on Image Processing. 2012. DOI : 10.1109/TIP.2011.2181525.
Doubly sparse models for multiple filter estimation in sparse echoic environmentsSignal Processing. 2012.
Clustering with Multi-Layer Graphs: A Spectral PerspectiveIEEE Transactions on Signal Processing. 2012. DOI : 10.1109/Tsp.2012.2212886.
A Laplacian pyramid scheme in graph signal processing2011
Hardware Implementation of an Omnidirectional Camera with Real-time 3D Imaging Capability3DTV 2011, Antalya, Turkey, May 16-18, 2011.
Chebyshev polynomial approximation for transductive learning on graphs2011
Guaranteed recovery of a low-rank and joint-sparse matrix from incomplete and noisy measurements (Abstract)2011. Workshop on Signal Processing with Adaptive Sparse Structured Representations (SPARS11), Edinburgh, UK, June 2011.
Audio-driven Nonlinear Video DiffusionIEEE Transactions on Image Processing. 2011.
Guaranteed recovery of a low-rank and joint-sparse matrix from incomplete and noisy measurementsSPARS11, Edinburgh, UK, June 27-29, 2011.
Methods for Clustering Multi-Layer Graphs in Mobile NetworksInterdisciplinary Workshop on Information and Decision in Social Networks, MIT, Cambridge, Massachusetts, USA, May 31-June 1, 2011.
Spread Spectrum for Universal Compressive Sampling2011. 4th Workshop on Signal Processing with Adaptive Sparse Structured Representations, Edinburgh, June 27-30, 2011. p. 50.
Implications for compressed sensing of a new sampling theorem on the sphere2011. 4th Workshop on Signal Processing with Adaptive Sparse Structured Representations, Edinburgh, Scotland, 27-30 June, 2011. p. 45.
Semi-Supervised Learning with Spectral Graph Wavelets2011. International Conference on Sampling Theory and Applications (SampTA), Singapore, May 2-6, 2011.
Image modeling with nonlocal spectral graph waveletsImage Processing and Analysing With Graphs: Theory and Practice; CRC Press, 2011.
Accelerated MR imaging with spread spectrum encoding2011. International Society for Magnetic Resonance in Medicine (ISMRM) conference, Montreal, May 7-13, 2011. p. 2808.
Classification via Incoherent SubspacesRejecta Mathematica. 2011.
Method and system for automatic objects localizationUS2014254875 ; US8749630 ; EP2386981 ; US2011279685 ; EP2386981 . 2010.
Estimating and Learning the Trajectory of Mobile Phones2010
Multichannel Compressed Sensing via Source Separation for Hyperspectral Images
Semi-supervised Extraction of Audio-Visual Sources2010
Stream Carving: an Adaptive Seam Carving Algorithm2010. International conference on Image Processing, Honk hong, September 26-29, 2010. p. 901-904. DOI : 10.1109/ICIP.2010.5653984.
Parallel Spread Spectrum MR Imaging2010
An optimal first-order solver for the TV-$L_1$ optical flow problem2010
Multichannel Compressed Sensing via Source Separation for Hyperspectral Images2010. Eusipco 2010, Aalborg, Denmark, 23-27 August, 2010.
Fast Structure from Motion for Planar Image Sequences2010. 2010 European Signal Processing Conference, Aalborg, August 23-27 2010.
Nonnegative Matrix Factorization and Spatial Covariance Model for Under-Determined Reverberant Audio Source Separation2010. 10th International Conference on Information Sciences, Signal Processing and their applications (ISSPA 2010), Kuala Lumpur , Malaysia, May 10-13, 2010.
Tracking and Structure from Motion2010
Wavelets on the SphereFour Short Courses on Harmonic Analysis; Boston: Birkhäuser, 2010. p. 131-174.
Compressed Sensing: When sparsity meets samplingOptical and Digital Image Processing - Fundamentals and Applications; Wiley-Blackwell, 2010.
Method for spatially scalable video codingUS7616824 ; US2006120614 . 2009.
A Variational Framework for Structure from Motion in Omnidirectional Image Sequences2009
Sparsity-driven People Localization Algorithm: Evaluation in Crowded Scenes Environments2009. IEEE International Workshop on Performance Evaluation of Tracking and Surveillance, Snowbird, Utah, December 7-10, 2009. DOI : 10.1109/PETS-WINTER.2009.5399487.
Compressed sensing for radio interferometry: prior-enhanced Basis Pursuit imaging techniques2009. SPARS'09, Saint-Malo, April 06-09, 2009.
A Master-Slave Approach to Detect and Match Objects Across Several Uncalibrated Moving Cameras2009
CMOS Compressed Imaging by Random ConvolutionIEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Taipei, Taiwan, 19 -24 April 2009.
Cosmic string detection from interferometric data of the microwave background radiation2009
The Panoptic Camera - Plenoptic interpolation in an omnidirectional polydioptric camera2009
A low complexity orthogonal matching pursuit for sparse signal approximation with shift-invariant dictionaries2009. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP09), Taipei, Taiwan, 2009. p. 3445-3448. DOI : 10.1109/ICASSP.2009.4960366.
TopicsData nowadays come in overwhelming volume. In order to cope with this deluge, we explore and use the benefits of geometry and symmetry in higher dimensional data. But volume is not the only problem: data models are also increasingly complex, mixing various components. We thus use sparse representations and dictionaries as dimensionality reduction tools to dig out information from complicated high-dimensional datasets and multichannel signals, or to model complex behaviours in more classical signals. Finally data can also be complex because they are collected on surfaces, or more generally manifolds, or because they are not scalar-valued. We thus explore extensions of Computational Harmonic Analysis in higher dimensions, in complex geometries, on graphs, networks or for non-scalar data.
Teaching & PhD
Electrical and Electronics Engineering