...

Video of Xi'an

FUSION 2017 TUTORIALS

Download Call For Tutorials

Click here  to download the call for Tutorials.

General Information

The main purpose of the tutorial program is to provide overviews of the state of the art in the particular subfields of Information Fusion. This program should be useful to newcomers to the Information Fusion community to learn about the research achievements of the particular areas, as well as to currently active Information Fusion researchers who may be interested in widening their areas of research interest.

Tutorials will be held on Monday, July 10, 2017 from 08:00 to 19:00.

Tutorials do not include lunch, but include coffee/tea breaks between tutorials.

Information for Tutorial Attendees

¡°Attendance Certificates¡± will be provided to attendees.

Information for Tutorial Presenters

  • Slides/Material: Microsoft PPT (*.ppt, *.pptx) or PDF files are preferred.

List of Tutorials of FUSION 2017

Preliminary Schedule

Schedule Room A Room B Room C Room D Room E Room F
08:00-09:30
Coffee Break (15mins)
09:45-11:15
40 years of tracking for radar systems: a cross disciplinary academic and industry point of view
T1-Farina, Chisci
Bayesian SLAM Algorithms
T3-Ristic
Analytic Combinatorics for Multi-target Tracking and higher level Fusion
T16-Streit, Efe
Multi-sensor Multi-target Tracker/Fusion Engine development and performance evaluation for realistic scenarios
T4-Kirubarajan
Robust Inference in emerging Networks: Challenges, Solutions and Directions
T11-Kailkhura, Vempathy, Varshney
Multi-sensor Fusion for Autonomous Vehicles
T9-Duraisamy, Yuan, Schwarz, Fritzche
11:15-11:30 Coffee Break
11:30-13:00
Lunch Break(1 hour)
14:00-15:30
Multi-target Tracking and Multi-sensor Fusion
T15-Bar-Shalom
Manoeuvring target Tracking: Overview and nonlinear filtering methods
T14-Li, Jilkov
Noise covariance matrices in state space models: Overview, algorithms and comparison of estimation methods
T6-Dunik, Straka
Object tracking, sensor fusion and situation awareness for self-driving vehicles: problems, solutions and directions
T5-Kirubarajan
Information Fusion and Decision making support with Belief Functions
T2-Dezert, Han
Space object tracking
T7-Mallick
15:30-15:45 Coffee Break
15:45-17:15
Coffee Break (15mins)
17:30-19:00
Trust Fusion and Bayesian Reasoning with Subjective Logic
T10-Zhang
Extended and Group Object Tracking
T8-Granstrom, Baum, Mihaylova
No Tutorial scheduled Computational Data Fusion and Analytics
T13-Das
No Tutorial scheduled Implementations of Random-finite-set based Multi-target filters
T17-Ba-Ngu Vo

T1: 40 years of tracking for radar systems: a cross-disciplinary academic and industry point of view

Description:
The talk will describe the intertwined R&D activities, along several decades, between academia and industry in conceiving and implementing - on live radar systems - tracking algorithms for targets in civilian as well as defence and security applications.
We trace back from the alpha-beta adaptive filter to modern random set filters passing thru Kalman algorithm (in its many embodiments), Multiple Model filters, Multiple Hypothesis Tracking, Joint Probabilistic Data Association, Particle filter for nonlinear non Gaussian models. Fusion from heterogeneous collocated as well as non-collocated sensor data are also mentioned. Applications to land, naval and airborne sensors are considered. Active as well as passive radar experiences are overviewed. The description will be a balanced look to both mathematical aspects as well as practical implementation issues including mitigation of real life system limitations.

Presenter: Alfonso Farina   and Luigi Chisci 

Further details and Presenter Biographies (PDF)

Back to top menu

T2: Information fusion and decision-making support with belief functions

Description:
The combination of qualitative or quantitative, imprecise, uncertain sources of evidences information is of main interest especially in the development of complex systems that have to deal with uncertainties and highly conflicting information/data with usually (but not necessarily) human interaction at some higher fusion level for efficient decision-making. This task is very difficult in general and many theories have been developed to deal with different kinds of uncertainties (randomness, fuzziness, epistemic nature, etc.), like probability theory, possibility theory, and belief functions theories (DST, TBM, DSmT), etc. In this tutorial we concentrate on the presentation of belief functions starting from Shafer¡¯s original idea up to the most recent developments proposed by Dezert and Smarandache. We will show through different examples the limits of DST and how the problems can be circumvented with DSmT and also the limitations of DSmT. The mathematical level of this tutorial and of the didactic examples chosen will be kept as simple as possible to be easily understood by all attendees, and specially those not familiar with belief functions. Aside the presentation of belief functions and their use with advanced techniques, we also present different methods for decision-making under uncertainty based on belief functions. Both aspects of decision-making (mono-criterion and multi-criteria) will be presented.

Presenter:Jean Dezert   and Deqiang Han 

Further details and Presenter Biographies (PDF)

Back to top menu

T3: Bayesian SLAM Algorithms

Description:
Simultaneous localisation andmapping (SLAM) is a statistical estimation problemin which a moving robot, equipped with a ranging sensor(s) and using odometry data, gradually creates a map of an unknown environment. The tutorial will explain the Bayesian SLAM algorithms (including the famous Gmapping algorithmavailable in Robot Operating System) in the context of both feature-based and occupancy gridmaps. The following topics will be covered:

  • A brief review of sequential Bayesian estimation (Kalman and particle filter)
  • Robot motionmodels
  • Types ofmaps and sensor models
  • Robot localisation (known map)
  • Robotic mapping (known robot pose)
  • Simultaneous localisation andmapping (SLAM)
  • Active (autonomous) SLAM

Presenter: Branko Ristic 

Further details and Presenter Biographies (PDF)

Back to top menu

T4: Multi-sensor Multi-target Tracker/Fusion Engine Development and Performance Evaluation for Realistic Scenarios

Description:
While numerous tracking and fusion algorithms are available in the literature, their implementation and application on real-world problems are still challenging. Since new algorithms continue to emerge, rapidly prototyping them, developing for production and evaluating them on real-world (or realistic) problems efficiently are also essential. In addition to reviewing state-of-the-art tracking algorithms, this tutorial will focus on a number of realistic multisensor-multitarget tracking problems, simulation of large-scale tracking scenarios, rapid prototyping, development of high performance real-time tracking/fusion software, and performance evaluation on realistic scenarios. A unified tracker framework that can handle a number of state-of-the-art algorithms like the Multiple Hypothesis Tracking (MHT) algorithm, Multiframe Assignment (MFA) tracker and the Joint (Integrated) Probabilistic Data Association (J(I)PDA) tracker is presented. Modules for preprocessing (e.g., coordinate transformations, clutter estimation, thresholding, registration), data association (e.g., 2-D assignment, multiframe assignment, k-best assignment), filtering (e.g., Kalman filter, Interacting Multiple Model (IMM) Estimator, Unscented Kalman filter) and postprocessing (e.g., prediction, classification) are discussed. Fusion software with different architectures is also presented. Integration of sensors like radar, ESA, angle-only, PCL and AIS/ADS-B is demonstrated. Side-by-side performance evaluation of multiple algorithms using more than 30 metrics on realistic large-scale tracking scenarios is presented. A hands-on approach with ISR360, which is an end-to-end real-time software suite for Intelligence, Surveillance and Reconnaissance, will be the cornerstone of this tutorial.
The topics will include Review of Bayesian state estimation, Multitarget tracking system architecture, Implementation of J(I)PDA/MHT/MFA trackers, Implementation of a multisensor fusion engine, Implementation of realistic simulators, Implementation of a track analytics engine, Performance evaluation of trackers (MOP/MOE), and Real-world examples.

Presenter: Thia Kirubarajan 

Further details and Presenter Biographies (PDF)

Back to top menu

T5: Object tracking, sensor fusion and situational awareness for assisted- and self-driving vehicles: Problems, solutions and directions

Description:
The automotive industry has been undergoing a major revolution in the last few years. Rapid advances have been made in assisted- and self-driving vehicles. As a result, vehicles have become more efficient and more automated. A number of automotive as well as technology companies are in the process of developing smart cars that can drive themselves. While totally self-driving cars are still in their infancy, some features like self-parking, proximity detection and lane identification have already made it into production in high-end vehicles. In spite of these recent developments, significantly more research is needed in order to perfect these nascent technologies and to make them ready for mass production. This provides the motivation for this tutorial.
In this tutorial, we aim to discuss a number of problems related to assisted- and self-driving vehicles, potential solutions and directions for research & development. The issues discussed in this tutorial will span multitarget tracking, multisensor fusion and situational awareness within the context of smart cars. We will also present some of the algorithms that are available in the open literature as well as those we have developed recently. In addition, we will also discuss related computational issues and sensor technologies. Finally, we will present some results on real data.

Presenter: Thia Kirubarajan 

Further details and Presenter Biographies (PDF)

Back to top menu

T6: Noise Covariance Matrices in State Space Models: Overview, Algorithms, and Comparison of Estimation Methods

Description:
Knowledge of a system model is a key prerequisite for many state estimation, signal processing, fault detection, and optimal control problems. The model is often designed to be consistent with random behaviour of the system quantities and properties of the measurements. While the deterministic part of the model often arises from mathematical modelling based on physical, chemical, or biological laws governing the behaviour of the system, the statistics of the stochastic part are often difficult to find by the modelling and have to be identified using the measured data. Incorrect description of the noise statistics may result in a significant worsening of estimation, signal processing, detection, or control quality or even in a failure of the underlying algorithms.
The tutorial introduces a more than six decades long history as well as recent advances and the state-of-the-art of the methods for estimation of the properties (or statistical description) of the stochastic part of the model with a special emphasis on the state-space model noise covariance matrices estimation. The tutorial covers all major groups of the noise statistics estimation methods, including the correlation methods, maximum likelihood methods, covariance matching methods, and the Bayesian methods. The methods are introduced in the unified framework highlighting their basic ideas, key properties, and assumptions. Algorithms of individual methods will be described and analysed to provide a basic understanding of their nature and similarities. Performance of the methods will also be compared using a numerical illustration.

Presenter: Jindrich Dunik   and  Ondrej Straka 

Further details and Presenter Biographies (PDF)

Back to top menu

T7: Space Object Tracking

Description:
Space objects (SOs) refer to Earth-orbiting satellites and space debris. In this tutorial, we provide and introduce orbital mechanics for space objects. Next, we describe equation of motion, force models (gravity, drag, solar radiation pressure (SRP)), and sensor measurement models. Finally, we present an overview of state-of-the-art filtering and tracking algorithms for SO tracking.
Space debris poses a serious threat to current and future space missions. At present, the number of SOs with size greater than one centimeter in low-Earth orbit (LEO) exceeds 300,000. LEO refers to orbits with altitudes up to 2000 km. Currently there are 1,419 (as of June 30, 2016) operating satellites from various countries and more than 22,000 SOs are tracked. More than 95% of the tracked SOs is space debris. Data on SOs from 1960 to 2000 show that the number of SOs is steadily increasing. NASA scientist Kessler predicted that continued production of space debris will eventually lead to a chain reaction where accidental collisions will increase exponentially; creating a debris shell in LEO that will render further operations in space impossible. This phenomenon is known as the Kessler syndrome. Therefore, tracking of SOs is of primary importance for future space missions.
The following material will be presented as part of this tutorial:

  • Overview of SO tracking: Current status of space objects and implications for future space program. Types of orbits - low-Earth orbit (LEO), mid-Earth orbit (MEO), geosynchronous orbit (GEO), and highly elliptical orbit (HEO). Precision orbit determination (OD) and OD for SOT.
  • Mathematical preliminaries coordinate frames and systems, and time systems
  • Introduction to orbit determination: Translational and rotational equations of motion for a SO.
  • Force models ? gravity, atmospheric drag, solar radiation pressure, and thrus
  • Sensors and measurement models: space surveillance network (SSN), radar and optical sensors, light?time corrections, aberration, and atmospheric refraction correction.
  • Two-body problem: Kepler¡¯s laws, orbital elements
  • Initial OD from range and angle measurements, OD from angle-only measurements, Gibbs algorithm, Lambert-Euler method, and Gauss method.
  • Review of nonlinear filtering algorithms for OD: Continuous-discrete filtering (CDF), weighted least squares (WLS) or differential correction (DC), extended Kalman filter (EKF), unscented Kalman filter (UKF), Gaussian sum filter (GSF).
  • Review of candidate algorithms for multitarget tracking: Centralized and distributed tracking, multiple hypothesis tracking (MHT), random finite set (RFS) based multitarget filtering algorithms, cardinalized probability hypothesis density (CPHD) and multi-Bernoulli (MeMBer) filters.

Presenter: Mahendra Mallick 

Further details and Presenter Biographies (PDF)

Back to top menu

T8: Extended and Group Object Tracking: Theory and Applications

Description:
Autonomous systems are an active area of research and technological development. These systems require intelligence and decision making, including intelligent sensing, data collection and processing, collision avoidance and control. Autonomous systems, especially autonomous cars need to be able to detect, recognise, classify and track objects of interest, including their location and size. In the light of autonomous systems his tutorial will focus on tracking of extended objects and group objects, i.e., object tracking using modern high resolution sensors that give multiple detections per object. State of the art theory will be introduced, and relevant real world applications will be shown where different object types, e.g., pedestrians, bicyclists, and cars, are tracked using different sensors such as lidar, radar, and camera.

Presenter: Karl Granstrom Marcus Baum   and  Lyudmila Mihalylova 

Further details and Presenter Biographies (PDF)

Back to top menu

T9: Multi-sensor Fusion for Autonomous Vehicles

Description:
This tutorial is focussed towards the stringent requirements, foundations, development and testing of multi sensor fusion algorithms meant for advanced driver assistance functions and likewise driverless applications in automotive vehicular systems. The audience would be provided with the .pdf materials used in the tutorial.
The audience can see the different representations of the surrounding environment as percepted by the heterogeneous environment perception sensors e.g. different radars, stereo camera and lidar. The relevant state estimation algorithms, sensor fusion frameworks and the evaluation procedures with reference ground truth are presented in detail. The audience can get an interesting glimpse of the data set obtained from a sensor configuration that would be used in the future Mercedes Benz autonomous vehicles.
More than one art of intelligent vehicular sensor fusion framework dealing with tracked objects i.e. track level fusion and raw sensor measurements i.e. measurement level fusion, with results obtained using several real world data sets that contains various static and dynamic targets would be presented in this tutorial. More focus would be provided on the important features of a low level fusion setup. Low level fusion and management of the different motion models of heterogeneous nature applied on sensors with different resolution is presented with examples. The impact of multiple model setup and a decorrelation procedure would be discussed in detail.
The interesting part of the tutorial is the coverage on the different challenging and important real time practical aspects of fusion such as tracking, data association, etc. related to an autonomous vehicular settings and requirements.

Presenter: Bharanidhar Duraisamy Ting Yuan Tilo Schwarz   and  Martin Fritzsche 

Further details and Presenter Biographies (PDF)

Back to top menu

T10: Trust Fusion and Bayesian Reasoning with Subjective Logic

Description:
This tutorial gives attendees a first-hand insight into the theory and application of subjective logic by the author and researcher who proposed and started developing this framework in 1997. In particular, it gives an introduction to subjective logic, and how it applies to reasoning under uncertainty, computational trust and trust fusion. Specific elements of the tutorial are:

1.Representation and interpretation of subjective opinions

  • a. Formal representation of binomial, multinomial and hyper opinions
  • b. Correspondence between subjective opinions and other relevant representations of trust such as binary logic propositions, probabilities, Dempster-Shafer belief functions,
  • c. Expressing opinions as PDFs (probability density functions) and qualitative measures

2.Algebraic operators of subjective logic

  • a. Operators for binomial opinions: transitivity, fusion, product, coproduct
  • b. Operators for multinomial opinions: conditional deduction and abduction, trust transitivity and fusion

3.Applications of subjective logic

  • a. Trust networks modelling and analysis
  • b. Subjective Bayesian reasoning modelling and analysis
  • c. Subjective networks based, on a combination of subjective Bayesian and trust networks

Presenter: Jie Zhang 

Further details and Presenter Biographies (PDF)

Back to top menu

T11:Robust Inference in Emerging Networks: Challenges, Solutions, and Directions

Description:
Distributed inference (detection, classification, localization, tracking) over networks has been an active area of research for almost a decade. Most of the work in the literature has been limited to the design and analysis of one-hop networks (parallel topology) with a fusion center (FC). However, in real world applications, the network is complex and dynamic, with multiple-hops to the FC or no FC at all (autonomous). Due to the complexity of real world networks, distributed inference solutions for such networks have been limitedly explored until recently. Also, another practical consideration is the unreliability of nodes in the network. These nodes can be unreliable due to a variety of reasons: noise, faults, and attacks, thus, resulting in corrupted data.
This tutorial will discuss the advances in the distributed inference literature by discussing fundamental limits of distributed inference with falsified data for several practical network topologies. An overview of optimal counter-measures to corrupted data (using the insights provided by these fundamental limits) is presented. For cases when problems are found to be challenging for practical cases (for example, NP-hard), several practical and simple to implement strategies are presented that overcome this challenge.
This tutorial will give a thorough introduction to various aspects of recent advances in robust distributed inference in ad-hoc networks. Any researcher, engineer, and graduate student who is interested in statistical inference can benefit from this tutorial. After taking this tutorial, they are expected to understand the principles behind distributed inference, learn advanced optimization techniques to solve inference problems in the presence of additional practical constraints, and get insights on inference in ad-hoc networks through practical examples. This tutorial will lay a foundation for students to solve real-world problems later.

Presenter: Bhavya Kailkhura Aditya Vempathy   and  Pramod Varshney 

Further details and Presenter Biographies (PDF)

Back to top menu

T13:Computational Data Fusion and Analytics

Description:
In this tutorial, I will present some techniques for fusion and analytics to process big centralized warehouse data, inherently distributed data, and data residing on the cloud. The broad range of computational artificial intelligence and machine learning techniques to be discussed will handle both structured transactional and sensor data as well as unstructured textual data such as human intelligence, emails, blogs, surveys, etc. Specifically, the tutorial will explore solving practical multi-sensor data fusion problems applying machine learning and artificial intelligence technologies.
As a background, this tutorial is intended to provide an account of both the cutting-edge and the most commonly used approaches to high-level data fusion and predictive and text analytics. The demos to be presented are in the areas of distributed search and situation assessment, information extraction and classification, and sentiment analyses.
Some of the tutorial materials are based on the following two books by the speaker: 1) Subrata Das. (2008). ¡°High-Level Data Fusion,¡± Artech House, Norwell, MA; and 2) Subrata Das. (2014). ¡°Computational Business Analytics,¡± Chapman & Hall/CRC Press.
Tutorial Topics include the following: High-Level Fusion, Descriptive and Predictive Analytics, Text Analytics, Machine Learning, Decision Support and Prescriptive Analytics, Machine Learning Models, Cloud Computing, Distributed Fusion, Hadoop and MapReduce, Natural Language Query, Big Data Query Processing, Graphical Probabilistic Models, Bayesian Belief Networks, Distributed Belief Propagation, Text Classification, Supervised and Unsupervised Classification, Information Extraction, Natural Language Processing.

Presenter: Subrata Das 

Further details and Presenter Biographies (PDF)

Back to top menu

T14: Maneuvering Target Tracking: Overview and nonlinear filtering methods

Description:
The principal challenges for tracking a maneuverable target are nonlinearity in both target motion and measurement models as well as the uncertainty in the pattern of target motion. This tutorial presents theoretical and algorithmic means available to meet these challenges. The overview part elucidates a well organized panorama of maneuvering target tracking. The other part presents an in-depth coverage of recent advances in nonlinear filtering for maneuvering target tracking, including some of the instructors¡¯ results and insights as well as better known methods. The tutorial highlights the underlying ideas and pros and cons of approaches and techniques as well as inter-relationships among them. It is an outgrowth of the instructors¡¯ ongoing comprehensive survey and several short courses of the same subject as well as a graduate course on target tracking taught at the Electrical Engineering Department of the University of New Orleans.

Presenter:Rong Li   and Vesselin Jilkov 

Further details and Presenter Biographies (PDF)

Back to top menu

T15: Multi-target Tracking and Multi-sensor Information Fusion

Description:
The main objective of this tutorial is to provide to the participants the latest state-of-the art techniques to estimate the states of multiple targets with multisensor information fusion. Tools for algorithm selection, design and evaluation will be presented. These form the basis of automated decision systems for advanced surveillance and targeting. The various information processing configurations for fusion are described, including the recently solved track-to-track fusion from heterogeneous sensors.

Presenter: Yaakov Bar-shalom 

Further details and Presenter Biographies (PDF)

Back to top menu

T16: Analytic Combinatorics for Multi-Object Tracking and Higher Level Fusion

Description:
This tutorial is designed to facilitate understanding of the classical theory of Analytic Combinatorics (AC) and how to apply it to problems in multi-object tracking and higher level data fusion. AC is an economical technique for encoding combinatorial problems?without information loss?into the derivatives of a generating function (GF). Exact Bayesian filters derived from the GF avoid the heavy accounting burden required by traditional enumeration methods. Although AC is an established mathematical field, it is not widely known in either the academic engineering community or the practicing data fusion/tracking community. This tutorial lays the groundwork for understanding the methods of AC, starting with the GF for the classical Bayes-Markov filter. From this cornerstone, we derive many established filters (e.g., PDA, JPDA, JIPDA, PHD, CPHD, MultiBernoulli, MHT) with simplicity, economy, and insight. We also show how to use the saddle point method (method of stationary phase) to find low complexity approximations of probability distributions and summary statistics.

Presenter:Roy Streit   and Murat Efe 

Further details and Presenter Biographies (PDF)

Back to top menu

T17: Implementations of Random-finite-set based Multi-target filters

Description:
In this tutorial, we show how the random finite set multi-target filters are implemented, and illustrate via Matlab how they work. Matlab code for these filters will be provided to all participants. The random finite set framework for multi-sensor multi-target tracking has attached considerable interest in recent years. It provides a unified perspective of multi-target tracking in a very intuitive manner by drawing direct parallels with the simpler problem of single-target tracking. This framework has led to the development of well-known multi-target filters such as the Probability Hypothesis Density (PHD), Cardinalized PHD (CPHD), Multi-Bernoulli filters and a recent advance, the Generalized Labeled Multi-Bernoulli filter. In particular, the tutorial will present the implementations of the (1) PHD filter (2) CPHD filter (3) Generalized Labeled Multi-Bernoulli filter¡ªa Bayes optimal multi-target tracker capable of tracking thousands of targets on a laptop. It is envisaged that participants will come away with sufficient know-how to implement and apply these algorithms in their work.

Presenter:Ba-Ngu Vo 

Further details and Presenter Biographies (PDF)

Back to top menu