Pierre Vandergheynst

Chattez avec Graph Search

Posez n’importe quelle question sur les cours, conférences, exercices, recherches, actualités, etc. de l’EPFL ou essayez les exemples de questions ci-dessous.

AVERTISSEMENT : Le chatbot Graph n'est pas programmé pour fournir des réponses explicites ou catégoriques à vos questions. Il transforme plutôt vos questions en demandes API qui sont distribuées aux différents services informatiques officiellement administrés par l'EPFL. Son but est uniquement de collecter et de recommander des références pertinentes à des contenus que vous pouvez explorer pour vous aider à répondre à vos questions.

This paper proposes a tree-based pursuit algorithm that efficiently trades off complexity and approximation performance for overcomplete signal expansions. Finding the sparsest representation of a signal using a redundant dictionary is, in general, a NP-Hard problem. Even sub-optimal algorithms such as Matching Pursuit remain highly complex. We propose a structuring strategy that can be applied to any redundant set of functions, and which basically groups similar atoms together. A measure of similarity base

Dense three-dimensional reconstruction of a scene from images is a very challenging task. In the structure from motion approach one of the key points is to compute depth maps which contain the distance of objects in the scene to a moving camera. Usually, this is achieved by finding correspondences in successive images and computing the distance by means of epipolar geometry. In this Master's thesis, a variational framework to solve the depth from motion problem for planar image sequences is proposed. Camera

Local Binary Descriptors (LBDs) are good at matching image parts, but how much information is actually carried? Surprisingly, this question is usually ignored and replaced by a comparison of matching performances. In this paper, we directly address it by t ...

This paper introduces Graph Convolutional Recurrent Network (GCRN), a deep learning model able to predict structured sequences of data. Precisely, GCRN is a generalization of classical recurrent neural networks (RNN) to data structured by an arbitrary graph. Such structured sequences can represent series of frames in videos, spatio-temporal measurements on a network of sensors, or random walks on a vocabulary graph for natural language modeling. The proposed model combines convolutional neural networks (CNN

This paper describes a novel framework for compressive sampling (CS) of multichannel signals that are highly dependent across the channels. In this work, we assume few number of sources are generating the multichannel observations based on a linear mixture ...

Radio interferometry probes astrophysical signals through incomplete and noisy Fourier measurements. The theory of compressed sensing demonstrates that such measurements may actually suffice for accurate reconstruction of sparse or compressible signals. We ...

Ultra-low power optimization is a challenging research topic in the de- sign of the sensor front-ends, especially in the area of Wireless Body Sensor Nodes (WBSN), where a limited amount of power and hardware resources are available. In this talk, I analyze the potential of the emerging compressed sensing (CS) paradigm for low-complexity and energy-efficient ECG sensing and data compression for storage or transmission, considering both software and hardware aspects. First, I discuss the power efficiency of

A novel model is presented to learn bimodally informative structures from audio-visual signals. The signal is represented as a sparse sum of audio- visual kernels. Each kernel is a bimodal function consisting of synchronous snippets of an audio waveform an ...

This work tackles the challenge of detecting and matching objects in scenes observed simultaneously by fixed and mobile cameras. No calibration between the cameras is needed, and no training data is used. A fully automated system is presented to detect if an object, observed by a fixed camera, is seen by a mobile camera and where it is localized in its image plane. Only the observations from the fixed camera are used. An object descriptor based on grids of region descriptors is used in a cascade manner. Fix

We consider the transductive learning problem when the labels belong to a continuous space. Through the use of spectral graph wavelets, we explore the benefits of multiresolution analysis on a graph constructed from the labeled and unlabeled data. The spec ...

The construction of a meaningful graph plays a crucial role in the success of many graph-based representations and algorithms for handling structured data, especially in the emerging field of graph signal processing. However, a meaningful graph is not alwa ...

In this work, we are interested in generalizing convolutional neural networks (CNNs) from low-dimensional regular grids, where image, video and speech are represented, to high-dimensional irregular domains, such as social networks, brain connectomes or wor ...

Graph-based methods for signal processing have shown promise for the analysis of data exhibiting irregular structure, such as those found in social, transportation, and sensor networks. Yet, though these systems are often dynamic, state-of-the-art methods ...

Polyamides, or peptides and proteins, are biomolecules that exist in a broad spectrum of size, structure, and function. Both structure and function are defined by the underlying sequence of amino acids, causing the polyamide to take three-dimensional conformations when in solution. Despite significant efforts and advances in function and conformation prediction, there remains a critical need for computational methods to accurately infer protein function from sequence and structure. Recent advancements in de

The development of more processing demanding applications on the Internet (video broadcasting) on one hand and the popularity of recent devices at the user level (digital cameras, wireless videophones, ...) on the other hand introduce challenges at several ...

We propose a novel method to automatically extract the audio-visual objects that are present in a scene. First, the synchrony between related events in audio and video channels is exploited to identify the possible locations of the sound sources. Video reg ...

This paper presents a novel approach to capture light field in camera arrays based on the tensor recovery framework. Light fields are acquired by a linear array of cameras with overlapping field of view. We represent the light fields in shape of a tensor t ...

The dictionary approach to signal and image processing has been massively investigated in the last two decades, proving very attractive for a wide range of applications. The effectiveness of dictionary-based methods, however, is strongly influenced by the choice of the set of basis functions. Moreover the structure of the dictionary is of paramount importance regarding efficient implementation and practical applications such as image coding. In this work, an overcomplete code for sparse representation of na

This paper presents a low rate progressive 3D mesh compression scheme for simple genus-zero 3D objects. The proposed scheme is based on signal representation using redundant expansions. 3D model representation is constructed using a Matching Pursuit algorithm, with an over-complete dictionary of atoms defined on a sphere. A specific dictionary construction is proposed, that is adapted to the characteristics of 3D models. The dictionary is built on both low-frequency atoms, and anisotropic refinement to capt

Many real world datasets subsume a linear or non-linear low-rank structure in a very low-dimensional space. Unfortunately, one often has very little or no information about the geometry of the space, resulting in a highly under-determined recovery problem. Under certain circumstances, state-of-the-art algorithms provide an exact recovery for linear low-rank structures but at the expense of highly inscalable algorithms which use nuclear norm. However, the case of non-linear structures remains unresolved. We

Pierre Vandergheynst

Chattez avec Graph Search

Tree-Based Pursuit: Algorithm and Properties

Tracking and Structure from Motion

Beyond Bits: Reconstructing Images from Local Binary Descriptors

Structured Sequence Modeling with Graph Convolutional Recurrent Networks

Distributed Compressed Sensing of Hyperspectral Images via Blind Source Separation

Compressed sensing imaging techniques for radio interferometry

Ultra low power design of hardware efficient CS-Based compression scheme in WBSN

Learning bimodal structure in audio-visual data

Object Detection and Matching in a Mixed Network of Fixed and Mobile Cameras

Semi-Supervised Learning with Spectral Graph Wavelets

Learning Laplacian Matrix in Smooth Graph Signal Representations

Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering

Towards Stationary Time-Vertex Signal Processing

Graph learning for capturing long-range dependencies in protein structures

Processor Enhancements for Media Streaming Applications

Audio-Visual Object Extraction using Graph Cuts

Light Field Tensor Recovery

Learning Structured Dictionaries for Image Representation

Progressive low bit rate coding of simple 3D objects with Matching Pursuit

Low-Rank Matrices on Graphs: Generalized Recovery & Applications