SLAM Course - 09 - Particle Filter Intro - Cyrill Stachniss

[@CyrillStachniss]

the grows is related to the number of samples that are needed to cover the high-dimensional state space with a sufficient number of samples to have a good approximation (not that the space is a Cartesian product). The grows can be smaller if the dimensions are correlated. I am not an expert in PHD filters but afaik the generic sequential MC is also computationally costly. I guess more efficient variants do approximations or assumptions... Not sure of that answers your question.

[@NoSwearld]

hey! can you please link me the lecture where you talked about motion and measurement model in detail? Thanks in advance!

[@chinyung90]

Hi! For a high dimensional space where we have multiple target, why is that the computational cost increase exponentially with increasing number of target? I mean how do we know that its exponentially increase? I have read many articles that mention abt this, but i dunno the reason behind. And also, how PHD filter can reduce the computational cost to linear just by approximating the first order moment of the posterior? Your help is much appreciated. :)

[@danchoithuthiet2009]

Thank you so much for your useful lecture! :-)

[@weselyong]

Thanks for sharing this !!

[@FarooqKifayat]

in the pseudo code for the particle filter algorithm shouldn't the second for loop be initialized as i=1:M. instead of m=1:M. ?

[@CyrillStachniss]

+Farooq Kifayatullah No, i is the index of the drawn particle and you do that m times.

[@FarooqKifayat]

Thks

[@멜리사-j3w]

Thank you very much :) This video was very helpful~

[@mojiheydari]

aweeeeesoooome

[@ChristyMedia-b4x]

how do i implement this in matlab ?

[@AmirMukhtar]

would be better not to see equations which confuse me. Instead listen to him and will understand more. idk why there is no easy translation and linkage of equations to what happens when we program a paticle filter. I can program and make others understand how particle filter works but can't explain in terms of equations.