-
Credal Valuation Networks for Machine Reasoning Under Uncertainty
Authors:
Branko Ristic,
Alessio Benavoli,
Sanjeev Arulampalam
Abstract:
Contemporary undertakings provide limitless opportunities for widespread application of machine reasoning and artificial intelligence in situations characterised by uncertainty, hostility and sheer volume of data. The paper develops a valuation network as a graphical system for higher-level fusion and reasoning under uncertainty in support of the human operators. Valuations, which are mathematical…
▽ More
Contemporary undertakings provide limitless opportunities for widespread application of machine reasoning and artificial intelligence in situations characterised by uncertainty, hostility and sheer volume of data. The paper develops a valuation network as a graphical system for higher-level fusion and reasoning under uncertainty in support of the human operators. Valuations, which are mathematical representation of (uncertain) knowledge and collected data, are expressed as credal sets, defined as coherent interval probabilities in the framework of imprecise probability theory. The basic operations with such credal sets, combination and marginalisation, are defined to satisfy the axioms of a valuation algebra. A practical implementation of the credal valuation network is discussed and its utility demonstrated on a small scale example.
△ Less
Submitted 4 August, 2022;
originally announced August 2022.
-
Machine Learning for automatic identification of new minor species
Authors:
Frederic Schmidt,
Guillaume Cruz Mermy,
Justin Erwin,
Severine Robert,
Lori Neary,
Ian R. Thomas,
Frank Daerden,
Bojan Ristic,
Manish R. Patel,
Giancarlo Bellucci,
Jose-Juan Lopez-Moreno,
Ann-Carine Vandaele
Abstract:
One of the main difficulties to analyze modern spectroscopic datasets is due to the large amount of data. For example, in atmospheric transmittance spectroscopy, the solar occultation channel (SO) of the NOMAD instrument onboard the ESA ExoMars2016 satellite called Trace Gas Orbiter (TGO) had produced $\sim$10 millions of spectra in 20000 acquisition sequences since the beginning of the mission in…
▽ More
One of the main difficulties to analyze modern spectroscopic datasets is due to the large amount of data. For example, in atmospheric transmittance spectroscopy, the solar occultation channel (SO) of the NOMAD instrument onboard the ESA ExoMars2016 satellite called Trace Gas Orbiter (TGO) had produced $\sim$10 millions of spectra in 20000 acquisition sequences since the beginning of the mission in April 2018 until 15 January 2020. Other datasets are even larger with $\sim$billions of spectra for OMEGA onboard Mars Express or CRISM onboard Mars Reconnaissance Orbiter. Usually, new lines are discovered after a long iterative process of model fitting and manual residual analysis. Here we propose a new method based on unsupervised machine learning, to automatically detect new minor species. Although precise quantification is out of scope, this tool can also be used to quickly summarize the dataset, by giving few endmembers ("source") and their abundances. We approximate the dataset non-linearity by a linear mixture of abundance and source spectra (endmembers). We used unsupervised source separation in form of non-negative matrix factorization to estimate those quantities. Several methods are tested on synthetic and simulation data. Our approach is dedicated to detect minor species spectra rather than precisely quantifying them. On synthetic example, this approach is able to detect chemical compounds present in form of 100 hidden spectra out of $10^4$, at 1.5 times the noise level. Results on simulated spectra of NOMAD-SO targeting CH$_{4}$ show that detection limits goes in the range of 100-500 ppt in favorable conditions. Results on real martian data from NOMAD-SO show that CO$_{2}$ and H$_{2}$O are present, as expected, but CH$_{4}$ is absent. Nevertheless, we confirm a set of new unexpected lines in the database, attributed by ACS instrument Team to the CO$_{2}$ magnetic dipole.
△ Less
Submitted 15 December, 2020;
originally announced December 2020.
-
Intermittent Information-Driven Search for Underwater Targets
Authors:
Branko Ristic,
Alex Skvortsov
Abstract:
The problem is area-restricted search for targets using an autonomous mobile sensing platform. Detection is imperfect: the probability of detection depends on the range to the target, while the probability of false detections is non-zero. The paper develops an intermittent information-driven search strategy, which combines fast and non-receptive displacement phase (ballistic phase) with a slow dis…
▽ More
The problem is area-restricted search for targets using an autonomous mobile sensing platform. Detection is imperfect: the probability of detection depends on the range to the target, while the probability of false detections is non-zero. The paper develops an intermittent information-driven search strategy, which combines fast and non-receptive displacement phase (ballistic phase) with a slow displacement sensing phase. Decisions where to move next, both in the ballistic phase and the slow displacement phase, are information-driven: they maximise the expected information gain. The paper demonstrates the efficiency of the proposed strategy in the context of a search for underwater targets: the searcher is an autonomous amphibious drone which can both fly and land or takeoff from the sea surface.
△ Less
Submitted 10 November, 2019;
originally announced November 2019.
-
Target tracking in the framework of possibility theory: The possibilistic Bernoulli filter
Authors:
Branko Ristic,
Jeremie Houssineau,
Sanjeev Arulampalam
Abstract:
The Bernoulli filter is a Bayes filter for joint detection and tracking of a target in the presence of false and miss detections. This paper presents a mathematical formulation of the Bernoulli filter in the framework of possibility theory, where uncertainty is represented using {\em possibility} functions, rather than {\em probability} distributions. Possibility functions model the uncertainty in…
▽ More
The Bernoulli filter is a Bayes filter for joint detection and tracking of a target in the presence of false and miss detections. This paper presents a mathematical formulation of the Bernoulli filter in the framework of possibility theory, where uncertainty is represented using {\em possibility} functions, rather than {\em probability} distributions. Possibility functions model the uncertainty in a non-additive manner, and have the capacity to deal with partial (incomplete) problem specification. Thus, the main advantage of the possibilistic Bernoulli filter, derived in this paper, is that it can operate even in the absence of precise measurement and/or dynamic model parameters. This feature of the proposed filter is demonstrated in the context of target tracking using multi-static Doppler shifts as measurements.
△ Less
Submitted 10 November, 2019;
originally announced November 2019.
-
Robust TMA using the possibility particle filter
Authors:
Branko Ristic,
Jeremie Houssineau,
Sanjeev Arulampalam
Abstract:
The problem is target motion analysis (TMA), where the objective is to estimate the state of a moving target from noise corrupted bearings-only measurements. The focus is on recursive TMA, traditionally solved using the Bayesian filters (e.g. the extended or unscented Kalman filters, particle filters). The TMA is a difficult problem and may cause the algorithms to diverge, especially when the meas…
▽ More
The problem is target motion analysis (TMA), where the objective is to estimate the state of a moving target from noise corrupted bearings-only measurements. The focus is on recursive TMA, traditionally solved using the Bayesian filters (e.g. the extended or unscented Kalman filters, particle filters). The TMA is a difficult problem and may cause the algorithms to diverge, especially when the measurement noise model is imperfect or mismatched. As a robust alternative to the Bayesian filters for TMA, we propose the recently introduced possibility filter. This filter is implemented in the sequential Monte Carlo framework, and referred to as the possibility particle filter. The paper demonstrates its superior performance against the standard particle filter in the presence of a model mismatch, and equal performance in the case of the exact model match.
△ Less
Submitted 31 May, 2018;
originally announced June 2018.
-
Autonomous search for a diffusive source in an unknown environment
Authors:
Branko Ristic,
Alex Skvortsov,
Andrew Walker
Abstract:
The paper presents an approach to olfactory search for a diffusive emitting source of tracer (e.g. aerosol, gas) in an environment with unknown map of randomly placed and shaped obstacles.
The measurements of tracer concentration are sporadic, noisy and without directional information. The search domain is discretised and modelled by a finite two-dimensional lattice. The links is the lattice rep…
▽ More
The paper presents an approach to olfactory search for a diffusive emitting source of tracer (e.g. aerosol, gas) in an environment with unknown map of randomly placed and shaped obstacles.
The measurements of tracer concentration are sporadic, noisy and without directional information. The search domain is discretised and modelled by a finite two-dimensional lattice. The links is the lattice represent the traversable paths for emitted particles and for the searcher. A missing link in the lattice indicates a blocked paths, due to the walls or obstacles. The searcher must simultaneously estimate the source parameters, the map of the search domain and its own location within the map. The solution is formulated in the sequential Bayesian framework and implemented as a Rao-Blackwellised particle filter with information-driven motion control. The numerical results demonstrate the concept and its performance.
△ Less
Submitted 6 June, 2013;
originally announced June 2013.
-
Performance Evaluation of Random Set Based Pedestrian Tracking Algorithms
Authors:
Branko Ristic,
Jamie Sherrah,
Ángel F. García-Fernández
Abstract:
The paper evaluates the error performance of three random finite set based multi-object trackers in the context of pedestrian video tracking. The evaluation is carried out using a publicly available video dataset of 4500 frames (town centre street) for which the ground truth is available. The input to all pedestrian tracking algorithms is an identical set of head and body detections, obtained usin…
▽ More
The paper evaluates the error performance of three random finite set based multi-object trackers in the context of pedestrian video tracking. The evaluation is carried out using a publicly available video dataset of 4500 frames (town centre street) for which the ground truth is available. The input to all pedestrian tracking algorithms is an identical set of head and body detections, obtained using the Histogram of Oriented Gradients (HOG) detector. The tracking error is measured using the recently proposed OSPA metric for tracks, adopted as the only known mathematically rigorous metric for measuring the distance between two sets of tracks. A comparative analysis is presented under various conditions.
△ Less
Submitted 25 October, 2012;
originally announced November 2012.
-
Modelling and Performance analysis of a Network of Chemical Sensors with Dynamic Collaboration
Authors:
Alex Skvortsov,
Branko Ristic
Abstract:
The problem of environmental monitoring using a wireless network of chemical sensors with a limited energy supply is considered. Since the conventional chemical sensors in active mode consume vast amounts of energy, an optimisation problem arises in the context of a balance between the energy consumption and the detection capabilities of such a network. A protocol based on "dynamic sensor collabor…
▽ More
The problem of environmental monitoring using a wireless network of chemical sensors with a limited energy supply is considered. Since the conventional chemical sensors in active mode consume vast amounts of energy, an optimisation problem arises in the context of a balance between the energy consumption and the detection capabilities of such a network. A protocol based on "dynamic sensor collaboration" is employed: in the absence of any pollutant, majority of sensors are in the sleep (passive) mode; a sensor is invoked (activated) by wake-up messages from its neighbors only when more information is required. The paper proposes a mathematical model of a network of chemical sensors using this protocol. The model provides valuable insights into the network behavior and near optimal capacity design (energy consumption against detection). An analytical model of the environment, using turbulent mixing to capture chaotic fluctuations, intermittency and non-homogeneity of the pollutant distribution, is employed in the study. A binary model of a chemical sensor is assumed (a device with threshold detection). The outcome of the study is a set of simple analytical tools for sensor network design, optimisation, and performance analysis.
△ Less
Submitted 4 November, 2011;
originally announced November 2011.
