SlideShare ist ein Scribd-Unternehmen logo
1 von 36
A FRIENDLY APPROACH TO PARTICLE FILTERS IN COMPUTER VISION Concepts, hints and examples 1 2/2/2011 Dr.-Ing. Marcos Nieto Doncel Investigador/Researcher mnieto@vicomtech.org
Outline Motivation Bayesianframework Samplingsolution Examples 2 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Motivation Forwhat? Obtainestimates of a recursive/dynamicsystem Let’sstay in computervisionapplications W H (x0,y0) 3 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Motivation Why? Deterministicapproach Probabilisticapproach vs 4 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Motivation How? Define yourtarget Define yourfunctions Select a type of filteradaptedto 1) and 2) Implement and run Optionally: Writeyourpaper and share : ) 5 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Bayesianfiltering Target: xk Itevolvesthrough time accordingtosomedynamics, properties, interaction, etc. W W H H (x0,y0) x0 y0 Prior / Dynamics / Transition… p(xk|xk-1) 6 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Bayesianfiltering Observations: z1:k Noisy, distorted, indirect Typically, differentdimensionaliy Likelihood / Observationmodel / Measurements… p(zk|xk) 7 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Bayesianfiltering Posterior distribution: p(xk|z1:k) Probability density function This is all you can expect to know Typicallywewant a point-estimate of thisdistribution At each time instant: x*k At theend 8 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Bayesianfiltering Time K-1 K K+1 p(zk|xk): Observation model zk-1 zk zk+1 Measurements (visible) xk-1 xk xk+1 States (hidden) p(xk|xk-1): Dynamic model 9 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Bayesianfiltering How? Prediction Use thedynamics, guessfutureaccordingto Correction Obtain a new observation, and applyBayes’ rule  Likelihood Prediction Posterior p(xk|xk-1) 10 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Particlefilters Solvethroughsampling! Letusapproximate posterior as a set of samples Samples / Particles / Hypotheses Weighted UnWeighted 11 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Particlefilters Understandingparticles Eachparticlerepresents a hypothesis Remember! wewilltypicallywantjustonepoint-estimate Bestparticle, mean particle, mode, median… W W H H (x0,y0) x0 y0 12 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Particlefilters Howtosample? Importancesampling MarkovChain Monte Carlo Gibbssampling Slicesampling … Howmanysamples? As much as requiredtotrackthe posterior! 13 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Particlefilters Sequentialimportancesampling (SIR) 14 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Particlefilters Sequentialimportancesampling (SIR) 15 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
SIR – example (I) Single object tracking 16 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
SIR – example (I) Linear-Gaussiandynamics Generate N samplesstartingfrompreviousstateaddingestimatedvelocity And someGaussiannoise Thenoisemakesthatsamples are different! 17 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
SIR - example (I) Likelihoodbasedonsegmentationorcolor histogram Evaluateeachpredictedsampleaccordingtothisvalue Likelihoodfunctionshouldreturnhighvaluesfor “good” hypotheses, and lowfor “bad” hypotheses 18 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
SIR – example (II) Eye-tracking Linear predictionwon’twork Theprojection of theeyemovementonthescreenisdifficulttopredict Define a combination of linear-Gaussian + uniform 19 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
20 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
21 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
SIR – example (II) ,[object Object],Try toadaptthedynamicmodelyou use withwhatyouthinkisthe real dynamics of whatyouwanttotrack Thismayimplyusing mixture models, accelerations, etc. Also, a goodlikelihoodmodelshouldincludesomecontinuousterm (like a uniform), in orderto cope withocclusions, so thatthetrackisnotlost 22 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
SIR Problems Requirednumber of samplesincreaseexponentiallywithproblemdimension Severalobjects/elements? Define a multimodal posterior and generatemultiplepoint-estimates Clusterparticles Increasestate vector dimension Variable number of objects? Addexternalhandler Includethenumber of objects as another variable toestimate 23 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Particlefilters MCMC More flexible Theproblem of dimensionissoftened Directlysamplefromthe posterior Researchers are focusing in MCMC Manyexcellentworksthatproposesolutionstomultipleobject, interaction, entering-exiting, number of samplesreduction, etc. 24 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Particlefilters MCMC Generate a Markovchain of samplesdirectlyfromthe posterior 25 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
MCMC Metropolis-Hastings Startsomehow Propose a movement Acceptwithprobabilityequaltothe ratio betweenproposedvalue and previousone Prob. = 1 ifproposedisbetterthanprevious Prob. = ratio ifnot Metropolis-Hastings allowsobtainingsamplesforanarbitrarydistributionbymaking a chainwhichacceptsorrejectsmovements 26 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
MCMC Each sample is a hypothesis of the state of all objects Multipleobjects State vector includingallthedimensions of allobjects Metropolis-Hastings: Generate a chain of N samples Foreachsample, use theinformation of allthesamples at theprevioustime instant After the chain is completed, we have the sample-basedapproximation of the posterior 27 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
MCMC Marginalizedproposalmoves Proposemovement of a single dimension at each new sample E.g.don’tpropose a move in alldimensionsforallobjects Choose a dimension randomly and update it Burn-in period Stop when stationary function is reached. Or when maximum number of samples is reached. x W y … … x W H x x W L 28 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
MCMC Interactionbetweenobjects MarkovRandomField (MRF) factor 29 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
MCMC Variable number of objects Add an external detector, and modify state size Reversible-Jump MCMC Define an Enter move (creates an object) Define an Exit move (removes an object) Define an Update move (updates existing objects) 30 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Discussion Whatshould I use? SIR MCMC Kalman? 31 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Discussion Ifdynamics and observation are linear, and withGaussiannoise Use Kalman, thisistheoptimumsolution Ifnot, considerusing a particlefilter 32 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Discussion SIR Use itif target dimensionislow (3-5) Use itifyou plan toparallelizeprocessing Rememberparticles are independentonefromanother Wouldrequireimportantdesignissuesfor Managingmultipleobjects Managing variable number of objects 33 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Discussion MCMC Use itifdimensionsincrease It can notbeparallelized Rememberthatparticlesform a chain, and eachonedependsonthepreviousone Adaptedtomultipleobjects MRF interactioniseasytoinsert Metropolis-Hastings can beefficientlyadaptedtomultipleobject 34 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
Summary Define your target Determine itsdynamics Define thelikelihood Select a filterthatadaptstotheproblem Implementit Runitcarefullyselectingtheappropriateparameters of yourfunctions, number of particles, etc. 35 Marcos Nieto, PhD     -     mnieto@vicomtech.org 2/2/2011
2/2/2011 36 Dr.-Ing. Marcos Nieto Doncel Investigador/Researcher mnieto@vicomtech.org http://marcosnieto.net/

Weitere ähnliche Inhalte

Ähnlich wie A FRIENDLY APPROACH TO PARTICLE FILTERS IN COMPUTER VISION

HAMS - Product and Prototype
HAMS - Product and PrototypeHAMS - Product and Prototype
HAMS - Product and PrototypeHAMSproject
 
How to break apart a monolithic system safely without destroying your team - ...
How to break apart a monolithic system safely without destroying your team - ...How to break apart a monolithic system safely without destroying your team - ...
How to break apart a monolithic system safely without destroying your team - ...Matthew Skelton
 
Teams and monoliths - Matthew Skelton - Velocity EU 2016
Teams and monoliths - Matthew Skelton - Velocity EU 2016Teams and monoliths - Matthew Skelton - Velocity EU 2016
Teams and monoliths - Matthew Skelton - Velocity EU 2016Skelton Thatcher Consulting Ltd
 
Human Behaviour Understanding using Top-View RGB-D Data
Human Behaviour Understanding using Top-View RGB-D DataHuman Behaviour Understanding using Top-View RGB-D Data
Human Behaviour Understanding using Top-View RGB-D DataDaniele Liciotti
 
Teams and monoliths - Matthew Skelton - Agile in the City Bristol 2016
Teams and monoliths - Matthew Skelton - Agile in the City Bristol 2016Teams and monoliths - Matthew Skelton - Agile in the City Bristol 2016
Teams and monoliths - Matthew Skelton - Agile in the City Bristol 2016Skelton Thatcher Consulting Ltd
 
Concept extraction from the web of things (3)
Concept extraction from the web of things (3)Concept extraction from the web of things (3)
Concept extraction from the web of things (3)Amélie Gyrard
 
A N E XTENSION OF P ROTÉGÉ FOR AN AUTOMA TIC F UZZY - O NTOLOGY BUILDING U...
A N  E XTENSION OF  P ROTÉGÉ FOR AN AUTOMA TIC  F UZZY - O NTOLOGY BUILDING U...A N  E XTENSION OF  P ROTÉGÉ FOR AN AUTOMA TIC  F UZZY - O NTOLOGY BUILDING U...
A N E XTENSION OF P ROTÉGÉ FOR AN AUTOMA TIC F UZZY - O NTOLOGY BUILDING U...ijcsit
 
Single layer perceptron in python
Single layer perceptron in pythonSingle layer perceptron in python
Single layer perceptron in pythonTahmina Zebin
 
Continuous Unsupervised Training of Deep Architectures
Continuous Unsupervised Training of Deep ArchitecturesContinuous Unsupervised Training of Deep Architectures
Continuous Unsupervised Training of Deep ArchitecturesVincenzo Lomonaco
 
Algorithmic Analysis to Video Object Tracking and Background Segmentation and...
Algorithmic Analysis to Video Object Tracking and Background Segmentation and...Algorithmic Analysis to Video Object Tracking and Background Segmentation and...
Algorithmic Analysis to Video Object Tracking and Background Segmentation and...Editor IJCATR
 
RuleML2015: Using PSL to Extend and Evaluate Event Ontologies
RuleML2015: Using PSL to Extend and Evaluate Event OntologiesRuleML2015: Using PSL to Extend and Evaluate Event Ontologies
RuleML2015: Using PSL to Extend and Evaluate Event OntologiesRuleML
 
John W. Vinti Particle Tracker Final Presentation
John W. Vinti Particle Tracker Final PresentationJohn W. Vinti Particle Tracker Final Presentation
John W. Vinti Particle Tracker Final PresentationJohn Vinti
 
Kantian Philosophy of Mathematics and Young Robots: Could a baby robot grow u...
Kantian Philosophy of Mathematics and Young Robots: Could a baby robot grow u...Kantian Philosophy of Mathematics and Young Robots: Could a baby robot grow u...
Kantian Philosophy of Mathematics and Young Robots: Could a baby robot grow u...Aaron Sloman
 
Theory and Applications of Monte Carlo Simulations by Chan V. (Ed.).pdf
Theory and Applications of Monte Carlo Simulations by Chan V. (Ed.).pdfTheory and Applications of Monte Carlo Simulations by Chan V. (Ed.).pdf
Theory and Applications of Monte Carlo Simulations by Chan V. (Ed.).pdfssuser941d48
 
Detection of fraud in financial blockchain-based transactions through big dat...
Detection of fraud in financial blockchain-based transactions through big dat...Detection of fraud in financial blockchain-based transactions through big dat...
Detection of fraud in financial blockchain-based transactions through big dat...CARLOS III UNIVERSITY OF MADRID
 
Survey Multiple Object Tracking Survey Paper
Survey Multiple Object Tracking Survey PaperSurvey Multiple Object Tracking Survey Paper
Survey Multiple Object Tracking Survey PaperPulasthiKarunaratne
 

Ähnlich wie A FRIENDLY APPROACH TO PARTICLE FILTERS IN COMPUTER VISION (20)

Why FPGA
Why FPGAWhy FPGA
Why FPGA
 
HAMS - Product and Prototype
HAMS - Product and PrototypeHAMS - Product and Prototype
HAMS - Product and Prototype
 
How to break apart a monolithic system safely without destroying your team - ...
How to break apart a monolithic system safely without destroying your team - ...How to break apart a monolithic system safely without destroying your team - ...
How to break apart a monolithic system safely without destroying your team - ...
 
Teams and monoliths - Matthew Skelton - Velocity EU 2016
Teams and monoliths - Matthew Skelton - Velocity EU 2016Teams and monoliths - Matthew Skelton - Velocity EU 2016
Teams and monoliths - Matthew Skelton - Velocity EU 2016
 
Human Behaviour Understanding using Top-View RGB-D Data
Human Behaviour Understanding using Top-View RGB-D DataHuman Behaviour Understanding using Top-View RGB-D Data
Human Behaviour Understanding using Top-View RGB-D Data
 
Teams and monoliths - Matthew Skelton - Agile in the City Bristol 2016
Teams and monoliths - Matthew Skelton - Agile in the City Bristol 2016Teams and monoliths - Matthew Skelton - Agile in the City Bristol 2016
Teams and monoliths - Matthew Skelton - Agile in the City Bristol 2016
 
Concept extraction from the web of things (3)
Concept extraction from the web of things (3)Concept extraction from the web of things (3)
Concept extraction from the web of things (3)
 
A N E XTENSION OF P ROTÉGÉ FOR AN AUTOMA TIC F UZZY - O NTOLOGY BUILDING U...
A N  E XTENSION OF  P ROTÉGÉ FOR AN AUTOMA TIC  F UZZY - O NTOLOGY BUILDING U...A N  E XTENSION OF  P ROTÉGÉ FOR AN AUTOMA TIC  F UZZY - O NTOLOGY BUILDING U...
A N E XTENSION OF P ROTÉGÉ FOR AN AUTOMA TIC F UZZY - O NTOLOGY BUILDING U...
 
TESTING
TESTINGTESTING
TESTING
 
Single layer perceptron in python
Single layer perceptron in pythonSingle layer perceptron in python
Single layer perceptron in python
 
Nanocomputing
NanocomputingNanocomputing
Nanocomputing
 
Multimodal Deep Learning
Multimodal Deep LearningMultimodal Deep Learning
Multimodal Deep Learning
 
Continuous Unsupervised Training of Deep Architectures
Continuous Unsupervised Training of Deep ArchitecturesContinuous Unsupervised Training of Deep Architectures
Continuous Unsupervised Training of Deep Architectures
 
Algorithmic Analysis to Video Object Tracking and Background Segmentation and...
Algorithmic Analysis to Video Object Tracking and Background Segmentation and...Algorithmic Analysis to Video Object Tracking and Background Segmentation and...
Algorithmic Analysis to Video Object Tracking and Background Segmentation and...
 
RuleML2015: Using PSL to Extend and Evaluate Event Ontologies
RuleML2015: Using PSL to Extend and Evaluate Event OntologiesRuleML2015: Using PSL to Extend and Evaluate Event Ontologies
RuleML2015: Using PSL to Extend and Evaluate Event Ontologies
 
John W. Vinti Particle Tracker Final Presentation
John W. Vinti Particle Tracker Final PresentationJohn W. Vinti Particle Tracker Final Presentation
John W. Vinti Particle Tracker Final Presentation
 
Kantian Philosophy of Mathematics and Young Robots: Could a baby robot grow u...
Kantian Philosophy of Mathematics and Young Robots: Could a baby robot grow u...Kantian Philosophy of Mathematics and Young Robots: Could a baby robot grow u...
Kantian Philosophy of Mathematics and Young Robots: Could a baby robot grow u...
 
Theory and Applications of Monte Carlo Simulations by Chan V. (Ed.).pdf
Theory and Applications of Monte Carlo Simulations by Chan V. (Ed.).pdfTheory and Applications of Monte Carlo Simulations by Chan V. (Ed.).pdf
Theory and Applications of Monte Carlo Simulations by Chan V. (Ed.).pdf
 
Detection of fraud in financial blockchain-based transactions through big dat...
Detection of fraud in financial blockchain-based transactions through big dat...Detection of fraud in financial blockchain-based transactions through big dat...
Detection of fraud in financial blockchain-based transactions through big dat...
 
Survey Multiple Object Tracking Survey Paper
Survey Multiple Object Tracking Survey PaperSurvey Multiple Object Tracking Survey Paper
Survey Multiple Object Tracking Survey Paper
 

Kürzlich hochgeladen

Scaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organizationScaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organizationRadu Cotescu
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptxHampshireHUG
 
Boost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfBoost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfsudhanshuwaghmare1
 
Exploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone ProcessorsExploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone Processorsdebabhi2
 
IAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsIAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsEnterprise Knowledge
 
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...apidays
 
Data Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonData Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonAnna Loughnan Colquhoun
 
How to convert PDF to text with Nanonets
How to convert PDF to text with NanonetsHow to convert PDF to text with Nanonets
How to convert PDF to text with Nanonetsnaman860154
 
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdfThe Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdfEnterprise Knowledge
 
What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?Antenna Manufacturer Coco
 
Driving Behavioral Change for Information Management through Data-Driven Gree...
Driving Behavioral Change for Information Management through Data-Driven Gree...Driving Behavioral Change for Information Management through Data-Driven Gree...
Driving Behavioral Change for Information Management through Data-Driven Gree...Enterprise Knowledge
 
Slack Application Development 101 Slides
Slack Application Development 101 SlidesSlack Application Development 101 Slides
Slack Application Development 101 Slidespraypatel2
 
[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdfhans926745
 
2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...Martijn de Jong
 
A Call to Action for Generative AI in 2024
A Call to Action for Generative AI in 2024A Call to Action for Generative AI in 2024
A Call to Action for Generative AI in 2024Results
 
A Year of the Servo Reboot: Where Are We Now?
A Year of the Servo Reboot: Where Are We Now?A Year of the Servo Reboot: Where Are We Now?
A Year of the Servo Reboot: Where Are We Now?Igalia
 
Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024The Digital Insurer
 
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking MenDelhi Call girls
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerThousandEyes
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsMaria Levchenko
 

Kürzlich hochgeladen (20)

Scaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organizationScaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organization
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
 
Boost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfBoost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdf
 
Exploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone ProcessorsExploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone Processors
 
IAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsIAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI Solutions
 
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
 
Data Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonData Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt Robison
 
How to convert PDF to text with Nanonets
How to convert PDF to text with NanonetsHow to convert PDF to text with Nanonets
How to convert PDF to text with Nanonets
 
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdfThe Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
 
What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?
 
Driving Behavioral Change for Information Management through Data-Driven Gree...
Driving Behavioral Change for Information Management through Data-Driven Gree...Driving Behavioral Change for Information Management through Data-Driven Gree...
Driving Behavioral Change for Information Management through Data-Driven Gree...
 
Slack Application Development 101 Slides
Slack Application Development 101 SlidesSlack Application Development 101 Slides
Slack Application Development 101 Slides
 
[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf
 
2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...
 
A Call to Action for Generative AI in 2024
A Call to Action for Generative AI in 2024A Call to Action for Generative AI in 2024
A Call to Action for Generative AI in 2024
 
A Year of the Servo Reboot: Where Are We Now?
A Year of the Servo Reboot: Where Are We Now?A Year of the Servo Reboot: Where Are We Now?
A Year of the Servo Reboot: Where Are We Now?
 
Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024
 
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected Worker
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed texts
 

A FRIENDLY APPROACH TO PARTICLE FILTERS IN COMPUTER VISION

  • 1. A FRIENDLY APPROACH TO PARTICLE FILTERS IN COMPUTER VISION Concepts, hints and examples 1 2/2/2011 Dr.-Ing. Marcos Nieto Doncel Investigador/Researcher mnieto@vicomtech.org
  • 2. Outline Motivation Bayesianframework Samplingsolution Examples 2 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 3. Motivation Forwhat? Obtainestimates of a recursive/dynamicsystem Let’sstay in computervisionapplications W H (x0,y0) 3 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 4. Motivation Why? Deterministicapproach Probabilisticapproach vs 4 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 5. Motivation How? Define yourtarget Define yourfunctions Select a type of filteradaptedto 1) and 2) Implement and run Optionally: Writeyourpaper and share : ) 5 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 6. Bayesianfiltering Target: xk Itevolvesthrough time accordingtosomedynamics, properties, interaction, etc. W W H H (x0,y0) x0 y0 Prior / Dynamics / Transition… p(xk|xk-1) 6 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 7. Bayesianfiltering Observations: z1:k Noisy, distorted, indirect Typically, differentdimensionaliy Likelihood / Observationmodel / Measurements… p(zk|xk) 7 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 8. Bayesianfiltering Posterior distribution: p(xk|z1:k) Probability density function This is all you can expect to know Typicallywewant a point-estimate of thisdistribution At each time instant: x*k At theend 8 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 9. Bayesianfiltering Time K-1 K K+1 p(zk|xk): Observation model zk-1 zk zk+1 Measurements (visible) xk-1 xk xk+1 States (hidden) p(xk|xk-1): Dynamic model 9 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 10. Bayesianfiltering How? Prediction Use thedynamics, guessfutureaccordingto Correction Obtain a new observation, and applyBayes’ rule Likelihood Prediction Posterior p(xk|xk-1) 10 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 11. Particlefilters Solvethroughsampling! Letusapproximate posterior as a set of samples Samples / Particles / Hypotheses Weighted UnWeighted 11 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 12. Particlefilters Understandingparticles Eachparticlerepresents a hypothesis Remember! wewilltypicallywantjustonepoint-estimate Bestparticle, mean particle, mode, median… W W H H (x0,y0) x0 y0 12 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 13. Particlefilters Howtosample? Importancesampling MarkovChain Monte Carlo Gibbssampling Slicesampling … Howmanysamples? As much as requiredtotrackthe posterior! 13 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 14. Particlefilters Sequentialimportancesampling (SIR) 14 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 15. Particlefilters Sequentialimportancesampling (SIR) 15 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 16. SIR – example (I) Single object tracking 16 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 17. SIR – example (I) Linear-Gaussiandynamics Generate N samplesstartingfrompreviousstateaddingestimatedvelocity And someGaussiannoise Thenoisemakesthatsamples are different! 17 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 18. SIR - example (I) Likelihoodbasedonsegmentationorcolor histogram Evaluateeachpredictedsampleaccordingtothisvalue Likelihoodfunctionshouldreturnhighvaluesfor “good” hypotheses, and lowfor “bad” hypotheses 18 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 19. SIR – example (II) Eye-tracking Linear predictionwon’twork Theprojection of theeyemovementonthescreenisdifficulttopredict Define a combination of linear-Gaussian + uniform 19 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 20. 20 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 21. 21 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 22.
  • 23. SIR Problems Requirednumber of samplesincreaseexponentiallywithproblemdimension Severalobjects/elements? Define a multimodal posterior and generatemultiplepoint-estimates Clusterparticles Increasestate vector dimension Variable number of objects? Addexternalhandler Includethenumber of objects as another variable toestimate 23 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 24. Particlefilters MCMC More flexible Theproblem of dimensionissoftened Directlysamplefromthe posterior Researchers are focusing in MCMC Manyexcellentworksthatproposesolutionstomultipleobject, interaction, entering-exiting, number of samplesreduction, etc. 24 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 25. Particlefilters MCMC Generate a Markovchain of samplesdirectlyfromthe posterior 25 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 26. MCMC Metropolis-Hastings Startsomehow Propose a movement Acceptwithprobabilityequaltothe ratio betweenproposedvalue and previousone Prob. = 1 ifproposedisbetterthanprevious Prob. = ratio ifnot Metropolis-Hastings allowsobtainingsamplesforanarbitrarydistributionbymaking a chainwhichacceptsorrejectsmovements 26 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 27. MCMC Each sample is a hypothesis of the state of all objects Multipleobjects State vector includingallthedimensions of allobjects Metropolis-Hastings: Generate a chain of N samples Foreachsample, use theinformation of allthesamples at theprevioustime instant After the chain is completed, we have the sample-basedapproximation of the posterior 27 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 28. MCMC Marginalizedproposalmoves Proposemovement of a single dimension at each new sample E.g.don’tpropose a move in alldimensionsforallobjects Choose a dimension randomly and update it Burn-in period Stop when stationary function is reached. Or when maximum number of samples is reached. x W y … … x W H x x W L 28 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 29. MCMC Interactionbetweenobjects MarkovRandomField (MRF) factor 29 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 30. MCMC Variable number of objects Add an external detector, and modify state size Reversible-Jump MCMC Define an Enter move (creates an object) Define an Exit move (removes an object) Define an Update move (updates existing objects) 30 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 31. Discussion Whatshould I use? SIR MCMC Kalman? 31 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 32. Discussion Ifdynamics and observation are linear, and withGaussiannoise Use Kalman, thisistheoptimumsolution Ifnot, considerusing a particlefilter 32 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 33. Discussion SIR Use itif target dimensionislow (3-5) Use itifyou plan toparallelizeprocessing Rememberparticles are independentonefromanother Wouldrequireimportantdesignissuesfor Managingmultipleobjects Managing variable number of objects 33 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 34. Discussion MCMC Use itifdimensionsincrease It can notbeparallelized Rememberthatparticlesform a chain, and eachonedependsonthepreviousone Adaptedtomultipleobjects MRF interactioniseasytoinsert Metropolis-Hastings can beefficientlyadaptedtomultipleobject 34 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 35. Summary Define your target Determine itsdynamics Define thelikelihood Select a filterthatadaptstotheproblem Implementit Runitcarefullyselectingtheappropriateparameters of yourfunctions, number of particles, etc. 35 Marcos Nieto, PhD - mnieto@vicomtech.org 2/2/2011
  • 36. 2/2/2011 36 Dr.-Ing. Marcos Nieto Doncel Investigador/Researcher mnieto@vicomtech.org http://marcosnieto.net/