The sparse coding is approximation/representation of signals with the minimum number of coefficients using an overcomplete set of elementary functions. This kind of approximations/ representations has found numerous applications in source separation, denoising, coding and compressed sensing. The adaptation of the sparse approximation framework to the coding problem of signals is investigated in this thesis. Open problems are the selection of appropriate models and their orders, coefficient quantization and sparse approximation method. Some of these questions are addressed in this thesis and novel methods developed. Because almost all recent communication and storage systems are digital, an easy method to compute quantized sparse approximations is introduced in the first part. The model selection problem is investigated next. The linear model can be adapted to better fit a given signal class. It can also be designed based on some a priori information ...

Sparse representations are intensively used in signal processing applications, like image coding, denoising, echo channels modeling, compression, classification and many others. Recent research has shown encouraging results when the sparse signals are created through the use of a learned dictionary. The current study focuses on finding new methods and algorithms, that have a parallel form where possible, for obtaining sparse representations of signals with improved dictionaries that lead to better performance in both representation error and execution time. We attack the general dictionary learning problem by first investigating and proposing new solutions for sparse representation stage and then moving on to the dictionary update stage where we propose a new parallel update strategy. Lastly, we study the effect of the representation algorithms on the dictionary update method. We also researched dictionary learning solutions where the dictionary has a specific form. ...

Many signal and image processing applications have benefited remarkably from the theory of sparse representations. In its classical form this theory models signal as having a sparse representation under a given dictionary -- this is referred to as the "Synthesis Model". In this work we focus on greedy methods for the problem of recovering a signal from a set of deteriorated linear measurements. We consider four different sparsity frameworks that extend the aforementioned synthesis model: (i) The cosparse analysis model; (ii) the signal space paradigm; (iii) the transform domain strategy; and (iv) the sparse Poisson noise model. Our algorithms of interest in the first part of the work are the greedy-like schemes: CoSaMP, subspace pursuit (SP), iterative hard thresholding (IHT) and hard thresholding pursuit (HTP). It has been shown for the synthesis model that these can achieve a stable recovery ...

Video signals are sequences of natural images, where images are often modeled as piecewise-smooth signals. Hence, video can be seen as a 3D piecewise-smooth signal made of piecewise-smooth regions that move through time. Based on the piecewise-smooth model and on related theoretical work on rate-distortion performance of wavelet and oracle based coding schemes, one can better analyze the appropriate coding strategies that adaptive video codecs need to implement in order to be efficient. Efficient video representations for coding purposes require the use of adaptive signal decompositions able to capture appropriately the structure and redundancy appearing in video signals. Adaptivity needs to be such that it allows for proper modeling of signals in order to represent these with the lowest possible coding cost. Video is a very structured signal with high geometric content. This includes temporal geometry (normally represented by motion ...

This thesis has two main parts. The first part deals with security and privacy analysis of biometric hashing. The second part introduces a method for fixed-length feature vector extraction and hash generation from fingerprint minutiae. The upsurge of interest in biometric systems has led to development of biometric template protection methods in order to overcome security and privacy problems. Biometric hashing produces a secure binary template by combining a personal secret key and the biometric of a person, which leads to a two factor authentication method. This dissertation analyzes biometric hashing both from a theoretical point of view and in regards to its practical application. For theoretical evaluation of biohashes, a systematic approach which uses estimated entropy based on degree of freedom of a binomial distribution is outlined. In addition, novel practical security and privacy attacks against face image hashing ...

Hyperspectral (HS) imaging, which consists of acquiring a same scene in several hundreds of contiguous spectral bands (a three dimensional data cube), has opened a new range of relevant applications, such as target detection [MS02], classification [C.-03] and spectral unmixing [BDPD+12]. However, while HS sensors provide abundant spectral information, their spatial resolution is generally more limited. Thus, fusing the HS image with other highly resolved images of the same scene, such as multispectral (MS) or panchromatic (PAN) images is an interesting problem. The problem of fusing a high spectral and low spatial resolution image with an auxiliary image of higher spatial but lower spectral resolution, also known as multi-resolution image fusion, has been explored for many years [AMV+11]. From an application point of view, this problem is also important as motivated by recent national programs, e.g., the Japanese next-generation space-borne ...

The unprecedented processing demand, posed by the explosion of big data, challenges researchers to design efficient and adaptive machine learning algorithms that do not require persistent retraining and avoid learning redundant information. Inspired from learning techniques of intelligent biological agents, identifying transferable knowledge across learning problems has been a significant research focus to improve machine learning algorithms. In this thesis, we address the challenges of knowledge transfer through embedding spaces that capture and store hierarchical knowledge. In the first part of the thesis, we focus on the problem of cross-domain knowledge transfer. We first address zero-shot image classification, where the goal is to identify images from unseen classes using semantic descriptions of these classes. We train two coupled dictionaries which align visual and semantic domains via an intermediate embedding space. We then extend this idea by training deep networks that ...

This thesis deals with developing improved modeling methods for speech and audio processing based on the recent developments in sparse signal representation. In particular, this work is motivated by the need to address some of the limitations of the well-known linear prediction (LP) based all-pole models currently applied in many modern speech and audio processing systems. In the first part of this thesis, we introduce \emph{Sparse Linear Prediction}, a set of speech processing tools created by introducing sparsity constraints into the LP framework. This approach defines predictors that look for a sparse residual rather than a minimum variance one, with direct applications to coding but also consistent with the speech production model of voiced speech, where the excitation of the all-pole filter is model as an impulse train. Introducing sparsity in the LP framework, will also bring to develop the ...

A novel and efficient methodology is presented for the conversion of stereo to multi-view video in order to address the 3D content requirements for the next generation 3D-TVs and auto-stereoscopic multi-view displays. There are two main algorithmic blocks in such a conversion system; stereo matching and virtual view rendering that enable extraction of 3D information from stereo video and synthesis of inexistent virtual views, respectively. In the intermediate steps of these functional blocks, a novel edge-preserving filter is proposed that recursively constructs connected support regions for each pixel among color-wise similar neighboring pixels. The proposed recursive update structure eliminates pre-defined window dependency of the conventional approaches, providing complete content adaptibility with quite low computational complexity. Based on extensive tests, it is observed that the proposed filtering technique yields better or competetive results against some leading techniques in the literature. The ...

Over-complete transforms have recently become the focus of a wide wealth of research in signal processing, machine learning, statistics and related fields. Their great modelling flexibility allows to find sparse representations and approximations of data that in turn prove to be very efficient in a wide range of applications. Sparse models express signals as linear combinations of a few basis functions called atoms taken from a so-called dictionary. Finding the optimal dictionary from a set of training signals of a given class is the objective of dictionary learning and the main focus of this thesis. The experimental evidence presented here focuses on the processing of audio signals, and the role of sparse algorithms in audio applications is accordingly highlighted. The first main contribution of this thesis is the development of a pitch-synchronous transform where the frame-by-frame analysis of audio data ...

This thesis interests in different methods of image compression combining both Bayesian aspects and ``sparse decomposition'' aspects. Two compression methods are in particular investigated. Transform coding, first, is addressed from a transform optimization point of view. The optimization is considered at two levels: in the spatial domain by adapting the support of the transform, and in the transform domain by selecting local bases among finite sets. The study of bases learned with an algorithm from the literature constitutes an introduction to a novel learning algorithm, which encourages the sparsity of the decompositions. Predictive coding is then addressed. Motivated by recent contributions based on sparse decompositions, we propose a novel Bayesian prediction algorithm based on mixtures of sparse decompositions. Finally, these works allowed to underline the interest of structuring the sparsity of the decompositions. For example, a weighting of the decomposition ...

Magnetic Resonance Imaging (MRI) constitutes a non-invasive medical imaging technique that allows the exploration of the inner anatomy, tissues, and physiological processes of the body. Among the different MRI applications, functional Magnetic Resonance Imaging (fMRI) has slowly become an essential tool for investigating the brain behavior and, nowadays, it plays a fundamental role in clinical and neurophysiological research. Due to its particular nature, specialized signal processing techniques are required in order to analyze the fMRI data properly. Among the various related techniques that have been developed over the years, the General Linear Model (GLM) is one of the most widely used approaches, and it usually appears as a default in many specialized software toolboxes for fMRI. On the other end, Blind Source Separation (BSS) methods constitute the most common alternative to GLM, especially when no prior information regarding the brain ...

Meningioma subtypes classification is a real world problem from the domain of histological image analysis that requires new methods for its resolution. Computerised histopathology presents a whole new set of problems and introduces new challenges in image classification. High intra-class variation and low inter-class differences in textures is often an issue in histological image analysis problems such as Meningioma subtypes classification. In this thesis, we present an adaptive wavelets based technique that adapts to the variation in the texture of meningioma samples and provides high classification accuracy results. The technique provides a mechanism for attaining an image representation consisting of various spatial frequency resolutions that represent the image and are referred to as subbands. Each subband provides different information pertaining the texture in the image sample. Our novel method, the Adaptive Discriminant Wavelet Packet Transform (ADWPT), provides a means for ...

This thesis develops new mathematical theory and presents novel recovery algorithms for discrete linear dynamical systems (LDS) with sparsity constraints on either control inputs or initial state. The recovery problems in this framework manifest as the problem of reconstructing one or more sparse signals from a set of noisy underdetermined linear measurements. The goal of our work is to design algorithms for sparse signal recovery which can exploit the underlying structure in the measurement matrix and the unknown sparse vectors, and to analyze the impact of these structures on the efficacy of the recovery. We answer three fundamental and interconnected questions on sparse signal recovery problems that arise in the context of LDS. First, what are necessary and sufficient conditions for the existence of a sparse solution? Second, given that a sparse solution exists, what are good low-complexity algorithms that ...

Multisensor data fusion technology combines data and information from multiple sensors to achieve improved accuracies and better inference about the environment than could be achieved by the use of a single sensor alone. In this dissertation, we propose parametric and nonparametric multisensor data fusion algorithms with a broad range of applications. Image registration is a vital first step in fusing sensor data. Among the wide range of registration techniques that have been developed for various applications, mutual information based registration algorithms have been accepted as one of the most accurate and robust methods. Inspired by the mutual information based approaches, we propose to use the joint R´enyi entropy as the dissimilarity metric between images. Since the R´enyi entropy of an image can be estimated with the length of the minimum spanning tree over the corresponding graph, the proposed information-theoretic registration ...