How is the discussion of HMM different from something like a Kalman filter to solve Mobile Robot Localization? In my understanding, our goal is to use the measurement feedback to improve our belief while respecting the system's own evolution dynamics. Sorry if the question is vague or incorrect.
This is the best hhm explanation I see.
How is the discussion of HMM different from something like a Kalman filter to solve Mobile Robot Localization? In my understanding, our goal is to use the measurement feedback to improve our belief while respecting the system's own evolution dynamics. Sorry if the question is vague or incorrect.
Kalman filter is a special case of HMM since it assumes all the relationships are linear.
Thanks Dr. Pascal. It is a good explanation. Could you add semi Markov models in the lecture series? A gentle request.