A5014128572- Image and Signal Denoising Methods
- Radio Astronomy Observations and Technology
- Sparse and Compressive Sensing Techniques
- Seismic Imaging and Inversion Techniques
- Geophysics and Gravity Measurements
- Cosmology and Gravitation Theories
- Galaxies: Formation, Evolution, Phenomena
- Statistical and numerical algorithms
- Gaussian Processes and Bayesian Inference
- Advanced Neuroimaging Techniques and Applications
- Mathematical Analysis and Transform Methods
- Medical Imaging Techniques and Applications
- Gamma-ray bursts and supernovae
- Advanced MRI Techniques and Applications
- Medical Image Segmentation Techniques
- Statistical Methods and Inference
- Optical measurement and interference techniques
- Soil Moisture and Remote Sensing
- Adaptive optics and wavefront sensing
- Antenna Design and Optimization
- Astronomy and Astrophysical Research
- Stellar, planetary, and galactic studies
- MRI in cancer diagnosis
- Microwave Imaging and Scattering Analysis
- Scientific Research and Discoveries
University College London
2015-2024
Turing Institute
2022-2024
The Alan Turing Institute
2023
British Library
2023
UCL Australia
2013-2015
École Polytechnique Fédérale de Lausanne
2008-2011
University of Puerto Rico-Mayaguez
2010
Mississippi State University
2010
University of Cambridge
2007-2008
Cavendish Hospital
2007
In a recent article series, the authors have promoted convex optimization algorithms for radio-interferometric imaging in framework of compressed sensing, which leverages sparsity regularization priors associated inverse problem and defines minimization image reconstruction. This approach was shown, theory through simulations simple discrete visibility setting, to potential outperform significantly CLEAN its evolutions. this work, we leverage versatility solving problems both handle...
We propose a novel algorithm for image reconstruction in radio interferometry. The ill-posed inverse problem associated with the incomplete Fourier sampling identified by visibility measurements is regularized assumption of average signal sparsity over representations multiple wavelet bases. algorithm, defined versatile framework convex optimization, dubbed Sparsity Averaging Reweighted Analysis (SARA). show through simulations that proposed approach outperforms state-of-the-art imaging...
The Large Synoptic Survey Telescope is designed to provide an unprecedented optical imaging dataset that will support investigations of our Solar System, Galaxy and Universe, across half the sky over ten years repeated observation. However, exactly how LSST observations be taken (the observing strategy or "cadence") not yet finalized. In this dynamically-evolving community white paper, we explore detailed performance anticipated science expected depend on small changes strategy. Using...
Automated photometric supernova classification has become an active area of research in recent years light current and upcoming imaging surveys such as the Dark Energy Survey (DES) Large Synoptic Telescope, given that spectroscopic confirmation type for all supernovae discovered will be impossible. Here, we develop a multi-faceted pipeline, combining existing new approaches. Our pipeline consists two stages: extracting descriptive features from curves using machine learning algorithm....
Abstract Next-generation surveys like the Legacy Survey of Space and Time (LSST) on Vera C. Rubin Observatory (Rubin) will generate orders magnitude more discoveries transients variable stars than previous surveys. To prepare for this data deluge, we developed Photometric LSST Astronomical Time-series Classification Challenge (PLAsTiCC), a competition that aimed to catalyze development robust classifiers under LSST-like conditions nonrepresentative training set large photometric test...
A new formalism is derived for the analysis and exact reconstruction of band-limited signals on sphere with directional wavelets. It represents an evolution wavelet developed by Antoine & Vandergheynst (1999) Wiaux et al. (2005). The translations wavelets at any point their proper rotations are still defined through continuous three-dimensional rotations. dilations directly in harmonic space a kernel dilation, which modification existing dilation. family factorized steerable functions...
We discuss a novel sparsity prior for compressive imaging in the context of theory compressed sensing with coherent redundant dictionaries, based on observation that natural images exhibit strong average over multiple frames. test our and associated algorithm, an analysis reweighted $\ell_1$ formulation, through extensive numerical simulations spread spectrum random Gaussian acquisition schemes. Our results show outperforms state-of-the-art priors promote single orthonormal basis or frame,...
In the context of next generation radio telescopes, like Square Kilometre Array, efficient processing large-scale datasets is extremely important. Convex optimisation tasks under compressive sensing framework have recently emerged and provide both enhanced image reconstruction quality scalability to increasingly larger data sets. We focus herein mainly on propose two new convex algorithmic structures able solve arising in radio-interferometric imaging. They rely proximal splitting...
Uncertainty quantification is a critical missing component in radio interferometric imaging that will only become increasingly important as the big-data era of interferometry emerges. Since requires solving high-dimensional, ill-posed inverse problem, uncertainty difficult but also to accurate scientific interpretation observations. Statistical sampling approaches perform Bayesian inference, like Markov Chain Monte Carlo (MCMC) sampling, can principle recover full posterior distribution...
Many areas of science and engineering encounter data defined on spherical manifolds. Modelling analysis often necessitates harmonic transforms, at high degrees, increasingly requires efficient computation gradients for machine learning or other differentiable programming tasks. We develop novel algorithmic structures accelerated generalised Fourier transforms the sphere S2 rotation group SO(3), i.e. Wigner respectively. present a recursive algorithm calculation d-functions that is both...
We develop a sampling scheme on the sphere that permits accurate computation of spherical harmonic transform and its inverse for signals band-limited at $L$ using only $L^2$ samples. obtain optimal number samples given by degrees freedom signal in space. The required our is factor two or four fewer than existing techniques, which require either $2L^2$ $4L^2$ note, however, we do not recover theorem sphere, where transforms are theoretically exact. Nevertheless, achieve high accuracy even...
Next-generation radio interferometers, such as the Square Kilometre Array (SKA), will revolutionise our understanding of universe through their unprecedented sensitivity and resolution. However, to realise these goals significant challenges in image data processing need be overcome. The standard methods interferometry for reconstructing images, CLEAN, have served community well over last few decades survived largely because they are pragmatic. produce reconstructed inter\-ferometric images...
We present the learned harmonic mean estimator with normalizing flows - a robust, scalable and flexible of Bayesian evidence for model comparison. Since is agnostic to sampling strategy simply requires posterior samples, it can be applied compute using any Markov chain Monte Carlo (MCMC) technique, including saved down MCMC chains, or variational inference approach. The was recently introduced, where machine learning techniques were developed learn suitable internal importance target...
We describe the construction of a spherical wavelet analysis through inverse stereographic projection Euclidean planar framework, introduced originally by Antoine and Vandergheynst developed further Wiaux et al. Fast algorithms for performing directional continuous on unit sphere are presented. The fast algorithm, based convolution algorithm Wandelt Gorski, provides saving O(sqrt(Npix)) over direct quadrature implementation Npix pixels sphere, allows one to perform 10^6 pixel map personal computer.
We develop an exact wavelet transform on the three-dimensional ball (i.e. solid sphere), which we name flaglet transform. For this purpose first construct radial half-line using damped Laguerre polynomials and a corresponding quadrature rule. Combined with spherical harmonic transform, approach leads to sampling theorem novel decomposition call Fourier-Laguerre relate new well-known Fourier-Bessel show that band-limitedness in basis is sufficient condition compute exactly. then through...
Uncertainty quantification is a critical missing component in radio interferometric imaging that will only become increasingly important as the big-data era of interferometry emerges. Statistical sampling approaches to perform Bayesian inference, like Markov Chain Monte Carlo (MCMC) sampling, can principle recover full posterior distribution image, from which uncertainties then be quantified. However, for massive data sizes, those anticipated Square Kilometre Array (SKA), it difficult if not...
Scale-discretised wavelets yield a directional wavelet framework on the sphere where signal can be probed not only in scale and position but also orientation. Furthermore, synthesised from its coefficients exactly, theory practice (to machine precision). are closely related to spherical needlets (both were developed independently at about same time) relax axisymmetric property of so that content probed. Needlets have been shown satisfy important quasi-exponential localisation asymptotic...
Abstract Bayesian model selection provides a powerful framework for objectively comparing models directly from observed data, without reference to ground truth data. However, requires the computation of marginal likelihood (model evidence), which is computationally challenging, prohibiting its use in many high-dimensional inverse problems. With imaging applications mind, this work we present proximal nested sampling methodology compare alternative that images inform decisions under...
A number of well-motivated extensions the $\ensuremath{\Lambda}\mathrm{CDM}$ concordance cosmological model postulate existence a population sources embedded in cosmic microwave background. One such example is signature bubble collisions which arise models eternal inflation. The most unambiguous way to test these scenarios evaluate full posterior probability distribution global parameters defining theory; however, direct evaluation computationally impractical on large datasets, as those...
We construct a directional spin wavelet framework on the sphere by generalising scalar scale-discretised transform to signals of arbitrary spin. The resulting is only defined natively that able probe intensity signals. Furthermore, wavelets support exact synthesis signal from its coefficients and satisfy excellent localisation uncorrelation properties. Consequently, are likely be use in wide range applications particular for analysis polarisation cosmic microwave background (CMB). develop...
The Photometric LSST Astronomical Time Series Classification Challenge (PLAsTiCC) is an open data challenge to classify simulated astronomical time-series in preparation for observations from the Large Synoptic Survey Telescope (LSST), which will achieve first light 2019 and commence its 10-year main survey 2022. revolutionize our understanding of changing sky, discovering measuring millions time-varying objects. In this challenge, we pose question: how well can objects sky that vary...
Curvelets are efficient to represent highly anisotropic signal content, such as a local linear and curvilinear structure. First-generation curvelets on the sphere, however, suffered from blocking artefacts. We present new second-generation curvelet transform, where scale-discretised constructed directly sphere. Scale-discretised exhibit parabolic scaling relation, well-localised in both spatial harmonic domains, support exact analysis synthesis of scalar spin signals, free fast algorithms...
Abstract Despite the technological advancements in Virtual Reality (VR), users are constantly combating feelings of nausea and disorientation, so-called cybersickness . Cybersickness symptoms cause severe discomfort hinder immersive VR experience. Here we investigated 360-degree head-mounted display VR. In traditional experiences, translational movement real world is not reflected virtual world, therefore self-motion information corroborated by matching visual vestibular cues, which may...
Abstract The Vera C. Rubin Observatory will increase the number of observed supernovae (SNe) by an order magnitude; however, it is impossible to spectroscopically confirm class for all SNe discovered. Thus, photometric classification crucial, but its accuracy depends on not-yet-finalized observing strategy Observatory’s Legacy Survey Space and Time (LSST). We quantitatively analyze impact LSST using simulated multiband light curves from Photometric Astronomical Time-Series Classification...