Least Squares - An Informal Introduction (Cyrill Stachniss)

[@Via.Dolorosa]

Thanks for this the neat way of your lecture, really appreciate your works in making those lectures available in a super-intelligent way.

[@neilpradhan1312]

Hello Professor, truely appreciate your effort to create such great content. It would be great if you could add link to the slides that you have presented in the video description.

[@vinhlai738]

Thanks a ton Professor! I'm diving the Multi-LiDARs project. Your videos really save my life a lot.

[@EcxaByte]

Really clear and well structured introduction to the topic, thanks a lot!

[@CyrillStachniss]

Thanks

[@hongkyulee9724]

Thank you for the nice lecture ! This lecture and slides are very helpful to me. I hope and want to be going to contribute the world like you :D

[@CyrillStachniss]

Thanks

[@martinschulze5399]

I found the explanation at 9:40 very confusing; Is x just a 3-d point of the input domain? why would you want to vary a point of the input domain instead of model parameters? the notation is a little bit ambiguous . I just assume x to be model parameters for now :)

[@sumitsarkar4517]

Here the problem is that of a state estimation problem rather than a model parameter estimation problem. Here the state is x (unknown) and the observation model is fi(x) (known and fixed). So the error vector is only function of x i.e. the state or as he explains the 3D coordinates (if state is assumed to be a postion in 3D world)

[@khaledhshmt1184]

wonderful! thanks professor.

[@jadtawil6143]

If x is in a the group of rotations (4 values of quaternion, for example). How do you constrain this algorithm so that the values of x iterated are valid quaternions?

[@javocremona]

In general, the optimization is performed on the associated Lie Algebra, see "A micro Lie theory
for state estimation in robotics" by Joan Sola et al.

[@amortalbeing]

thanks a lot really appreciate it

[@simplyclarified6948]

Clear story!