We study the spatiotemporal sampling of a diffusion field generated by K point sources, aiming to fully reconstruct the unknown initial field distribution from the sample measurements. The sampling operator in our problem can be described by a matrix derived from the diffusion model. We analyze the important properties of the sampling matrices, leading to precise bounds on the spatial and temporal sampling densities under which perfect field reconstruction is feasible. Moreover, our analysis indicates that it is possible to compensate linearly for insufficient spatial sampling densities by oversampling in time. Numerical simulations on initial field reconstruction under different spatiotemporal sampling densities confirm our theoretical results.

We consider the problem of estimating room geometry from the acoustic room impulse response (RIR). Existing approaches addressing this problem exploit the knowledge of multiple RIRs. In contrast, we are interested in reconstructing the room geometry from a single RIR — a 1–D function of time. We discuss the uniqueness of the mapping between the geometry of a planar polygonal room and a single RIR. In addition to this theoretical analysis, we also propose an algorithm that performs the "blindfolded" room estimation. Furthermore, the derived results are used to construct an algorithm for localization in a known room using only a single RIR. Verification of the theoretical developments with numerical simulations is given before concluding the paper.

We propose a novel algorithm for sparse system identification in the frequency domain. Key to our result is the observation that the Fourier transform of the sparse impulse response is a simple sum of complex exponentials, whose parameters can be efficiently determined from only a narrow frequency band. From this perspective, we present a sub-Nyquist sampling scheme, and show that the original continuous-time system can be learned by considering an equivalent low-rate discrete system. The impulse response of that discrete system can then be adaptively obtained by a novel frequency-domain LMS filter, which exploits the parametric structure of the model. Numerical experiments confirm the effectiveness of the proposed scheme for sparse system identification tasks.

We study a multichannel sampling scheme, where different channels observe scaled and shifted versions of a common bandlimited signal. The channel gains and offsets are unknown a priori, and each channel samples at sub-Nyquist rates. This setup appears in many practical signal processing applications, including time-interleaved ADC with timing skews, unsynchronized distributed sampling in sensor networks, and superresolution imaging. In this paper, we propose a new al- gorithm to efficiently estimate the unknown channel gains and offsets. Key to our algorithm is a novel linearization technique, which converts a system of trigonometric polynomial equations of the unknown parameters to an overparameterized linear system. The computation steps of the proposed algorithm boil down to forming a fixed data matrix from the discrete-time Fourier transforms of the observed channel samples and computing the singular value decomposition of that matrix. Numerical simulations verify the effectiveness, efficiency, and robustness of the proposed algorithm in the presence of noise. In the high SNR regime (40 dB and above), the proposed algorithm significantly outperforms a previous method in the literature in terms of estimation accuracy, at the same time being three orders of magnitudes faster.

We propose a Fourier analytical approach to the problems of alias-free sampling and critical sampling. Central to this approach are two Fourier conditions linking the above sam- pling criteria with the Fourier transform of the indicator function defined on the underlying frequency support. We present several examples to demonstrate the usefulness of the proposed Fourier conditions in the design of critically sampled multidimensional filter banks. In particular, we show that it is impossible to implement any cone-shaped fre- quency partitioning by a nonredundant filter bank, except for the 2-D case.

We consider the task of recovering correlated vectors at a central decoder based on fixed linear measurements ob- tained by distributed sensors. Two different scenarios are considered: In the case of universal reconstruction, we look for a sensing and recovery mechanism that works for all possible signals, whereas in the case of almost sure recon- struction, we allow to have a small set (with measure zero) of unrecoverable signals. We provide achievability bounds on the number of samples needed for both scenarios. The bounds show that only in the almost sure setup can we ef- fectively exploit the signal correlations to achieve effective gains in sampling efficiency. In addition, we propose an efficient and robust distributed sensing and reconstruction algorithm based on annihilating filters.

Digital camera sensors are inherently sensitive to the near- infrared (NIR) part of the light spectrum. In this paper, we propose a general design for color filter arrays that allow the joint capture of visible/NIR images using a single sensor. We pose the CFA design as a novel spatial domain optimization problem, and provide an efficient iterative procedure that finds (locally) optimal solutions. Numerical experiments confirm the effectiveness of the proposed CFA design, which can simultane- ously capture high quality visible and NIR image pairs.

Most digital cameras employ a spatial subsampling process, implemented as a color filter array (CFA), to capture color images. The choice of CFA patterns has a great impact on the performance of subsequent reconstruction (demosaicking) algorithms. In this work, we propose a quantitative theory for optimal CFA design. We view the CFA sampling process as an encoding (low-dimensional approximation) operation and, correspondingly, demosaicking as the best decoding (reconstruction) operation. Finding the optimal CFA is thus equivalent to finding the optimal approximation scheme for the original signals with minimum information loss. We present several quantitative conditions for optimal CFA design, and propose an efficient computational procedure to search for the best CFAs that satisfy these conditions. Numerical experiments show that the optimal CFA patterns designed from the proposed procedure can effectively retain the information of the original full-color images. In particular, with the designed CFA patterns, high quality demosaicking can be achieved by using simple and efficient linear filtering operations in the polyphase domain. The visual qualities of the reconstructed images are competitive to those obtained by the state-of-the-art adaptive demosaicking algorithms based on the Bayer pattern.

We study the spatio-temporal sampling of a diffusion field driven by K unknown instantaneous source distributions. Exploiting the spatio-temporal correlation offered by the diffusion model, we show that it is possible to compensate for insufficient spatial sampling densities (i.e. sub-Nyquist sampling) by increasing the temporal sampling rate, as long as their product remains roughly a constant. Combining a distributed sparse sampling scheme and an adaptive feedback mechanism, the proposed sampling algorithm can accurately and efficiently estimate the unknown sources and reconstruct the field. The total number of samples to be transmitted through the network is roughly equal to the number of degrees of freedom of the field, plus some additional costs for in-network averaging.

We study the spatial-temporal sampling of a linear diffusion field, and show that it is possible to compensate for insufficient spatial sampling densities by oversampling in time. Our work is motivated by the following issue often encountered in sensor network sampling, namely increasing the temporal sampling density is often easier and less expensive than increasing the spatial sampling density of the network. For the case of sampling a diffusion field, we show that, to achieve trade-off between spatial and temporal sampling, the spatial arrangement of the sensors must satisfy certain conditions. We provide in this paper the precise relationships between the achievable reduction of spatial sampling density, the required temporal oversampling rate, the spatial arrangement of the sensors, and the bound for the condition numbers of the resulting sampling and reconstruction procedures.

We consider the task of recovering correlated vectors at a central decoder based on fixed linear measurements obtained by distributed sensors. A general formulation of the problem is proposed, under both a universal and an almost sure reconstruction requirement. We then study a specific correlation model which involves a filter that is sparse in the time domain. While this sparsity assumption does not allow reducing the description cost in the universal case, we show that large gains can be achieved in the almost sure scenario by means of a novel distributed scheme based on annihilating filters. The robustness of the proposed method is also investigated.

Color image demosaicking is a key process in the digital imaging pipeline. In this paper, we present a rigorous treatment of a classical demosaicking algorithm based on alternating projections (AP). Since its publication, the AP algorithm has been widely cited and served as a benchmark in a flurry of papers in the demosaicking literature. Despite its impressive performances, a relative weakness of the AP algorithm is its high computational complexity. In our work, we provide a rigorous analysis of the convergence of the AP algorithm based on the concept of contraction mapping. Furthermore, we propose an efficient noniterative implementation of the AP algorithm in the polyphase domain. Numerical experiments show that the proposed noniterative implementation achieves the same results obtained by the original AP algorithm at convergence, but is about an order of magnitude faster than the latter.

SensorScope is a collaborative project between network, signal processing, and environmental researchers that aims at providing a cheap and out-of-the-box environmental monitoring system based on a wireless sensor network. It has been successfully used in a number of deployments to gather hundreds of megabytes of environmental data. With data gathering techniques well mastered, the efficient processing of the huge amounts of the acquired information to allow for useful exploitation has become an increasingly important issue. In this paper, we present a number of challenging and relevant signal processing tasks that arise from the SensorScope project. We believe the resolution of these problems will benefit from a better understanding of the underlying physical processes. We show an example to demonstrate how physical correlations between different sensing modalities can help reduce the sampling rate.

The search for alias-free sampling lattices for a given frequency support, in particular those lattices achieving minimum sam- pling densities, is a fundamental issue in various applications of signal and image processing. In this paper, we propose an efficient computational procedure to find all alias-free integral sampling lattices for a given frequency support with minimum sampling density. Central to this algorithm is a novel condition linking the alias-free sampling with the Fourier transform of the indicator function defined on the frequency support. We study the computation of these Fourier transforms based on the diver- gence theorem, and propose a simple closed-form formula for a fairly general class of support regions consisting of arbitrary N -dimensional polytopes, with polygons in 2-D and polyhedra in 3-D as special cases. The proposed algorithm can be useful in a variety of applications involving the design of efficient ac- quisition schemes for multidimensional bandlimited signals.

With the ever increasing computational power of modern day processors, it has become feasible to use more robust and computationally complex algorithms that increase the resolution of images without distorting edges and contours. We present a novel image interpolation algorithm that uses the new contourlet transform to improve the regularity of object boundaries in the generated images. By using a simple wavelet-based linear interpolation scheme as our initial estimate, we use an iterative projection process based on two constraints to drive our solution towards an improved high-resolution image. Our experimental results show that our new algorithm significantly outperforms linear interpolation in subjective quality, and in most cases, in terms of PSNR as well.

Motion estimation is a common ingredient in many state-of- the-art video processing algorithms, serving as an effective way to capture the spatial-temporal correlation in video signals. However, the robustness of motion estimation often suffers from problems such as ambiguities of motion trajectory (i.e. the aperture problem) and illumination variances. In this paper, we explore a new framework for video processing based on the recently proposed surfacelet transform. Instead of containing an explicit motion estimation step, the surfacelet transform provides a motion-selective subband decomposition for video signals. We demonstrate the potential of this new technique in a video denoising application.