Visual Tracking

18
Visual Tracking CMPUT 615 Nilanjan Ray

description

Visual Tracking. CMPUT 615 Nilanjan Ray. What is Visual Tracking. Following objects through image sequences or videos Sometimes we need to track a single object, sometimes a number of them Sometimes we just track the object centroid , sometimes entire object boundary (shape). - PowerPoint PPT Presentation

Transcript of Visual Tracking

Page 1: Visual Tracking

Visual Tracking

CMPUT 615Nilanjan Ray

Page 2: Visual Tracking

What is Visual Tracking

• Following objects through image sequences or videos

• Sometimes we need to track a single object, sometimes a number of them

• Sometimes we just track the object centroid, sometimes entire object boundary (shape)

Page 3: Visual Tracking

Theoretical Foundation

• Visual tracking is a “state” estimation problem• Bayesian inference is at the heart of visual

tracking; it is called sequential Bayesian estimation

• We form the posterior probability of the state, given all evidence or measurements up to the current time point

• Inference is performed from the posterior density

Page 4: Visual Tracking

Setting The Stage

Some notations:

Xt: unknown state we want to estimate at time point t; e.g., object centroid

Zt: Measurement/observation made at time point t; e.g., image intensities

The sequential estimation model assumes that we know three probabilitydensities:

p(X0): The initial state density

p(Xt|Xt-1): State transition density or motion model

p(Zt|Xt): Measurement/observation/likelihood density

Page 5: Visual Tracking

Sequential Bayesian Estimation(AKA Sequential Bayesian Filtering)

• We want to recursively estimate the state Xt given the observations Z1:t = {Z1, Z2, …, Zt}

Page 6: Visual Tracking

Sequential Bayesian Estimation…

Filter:

Bayes’ Rule:

Current posterior

Previous posteriorPrediction

Likelihood/observation density

Page 7: Visual Tracking

Bayes’ Rule Derivation

Conditional probability ruleMarginal density rule

Also, because measurement Zt is conditionally independent on the current state Xt:

So, we have the sequential Bayes’ rule:

Page 8: Visual Tracking

Filter Derivation

Rule of marginal density Rule of conditional probability

Also, note that Xt is conditionally independent on Xt-1 (Markovianity), so:

Thus we have the filter rule:

Page 9: Visual Tracking

Important Assumptions

• Observation is conditionally independent on the current state

• Current state is conditionally independent on the immediate previous state

Page 10: Visual Tracking

Computation

• Theory is all good, however we need to show people that it works in practice…

• We will study Particle filter, the framework that can compute the recursive state estimation, i.e., sequential Bayesian estimation

• We will also study Kalman filter, a popular sequential state estimation technique with some more assumptions

Page 11: Visual Tracking

What is a Particle Filter?Let the particles represent the previous density

So, the filter step is now:

And the Bayes’ rule is now:

We need to generate the current particle set from p(Xt|Zt):

Particle filter

Page 12: Visual Tracking

Factored Sampling

Let h(x) = f(x)g(x) is a product of two functions, where say, g(x) is a density andf(x) is another non-negative function

Factored sampling says that to represent h(x) non-parametrically by a set of particles, generate samples from g(x) and assign weights by f(x)

i.e., {(s1 , w1), …, (sn, wn)}, where si are generated from g(x) and wi = f(si).

This is closely related to another sampling method called importance sampling.

Page 13: Visual Tracking

Conditional Density Propagation (CONDENSATION)

This a product of two functions: (1)

and (2)

Following the principle of factored sampling, CONDENSATION generates samples from (1)And assigns weights using (2)

Page 14: Visual Tracking

Samples From a Mixture Density

Notice that is a mixture density

To generate samples from the mixture density these two steps are followed:

Page 15: Visual Tracking

CONDENSATION Algorithm

Page 16: Visual Tracking

How to estimate the state?

• OK, we generated samples, what do we do with them: estimate the current state:

h is any function of the state, for example when h(x) = x, we are performing state estimation

Page 17: Visual Tracking

Other PFs

• To date lots of particle filters have been proposed:– Sequential importance re-sampling (SIR)– Auxiliary particle filter (APF)– Likelihood particle filter– Rao-Blackwellized particle filter– A ton others

• A leading researcher in PF : Arnoud Doucet

Page 18: Visual Tracking

Some Points to Ponder about PF

• The good point about PF is that it can handle very general likelihood and motion models

• PF inherits a serious shortcoming from non-parametric density representation – curse of dimensionality – when the state space x is large, for example large multiple number of objects etc.