Inizio della pagina -
Logo DISCO
|
Visita la Versione ad elevata leggibilità
|
Vai al Contenuto della pagina
|
Vai alla Fine dei contenuti
|
Vai al Menu Principale
|
Vai alla Barra di navigazione (sei in)
|
Vai al Menu di navigazione (albero)
|
Vai alla Lista dei comandi
|
Vai alla Lista degli approfondimenti
|
Vai al Menu inferiore
|
Logo Ateneo
   
Bayes Filtering with Application to Robotic Perception
Docente

Prof. D.G Sorrenti 

Pedro Pinìes (Univ. Zaragoza, Es) will give the classes on EKF vision-based SLAM as well as the corresponding labs.

Data e luogo

 

Introduction to Bayes Filtering

  • Monday May 17th, 2010; 09:30 - 12:30; U14-02 aka T023
Basics of probabilities
Basics of filtering
Gaussian Filtering (Kalman)
  • Wednesday May 19th, 2010; 09:30 - 12:30; U14-03 aka T014
Extended Kalman Filter
Jacobian-based propagation of uncertainty
EKF for parameter estimation with an implicit output transform
Unscended Kalman Filter
  • Friday May 21th, 2010; 09:30 - 12:30; Sala Seminari
Information Form Gaussian Filtering
Non parametric filtering
  • Wednesday May 26th, 2010; 09:30 - 12:30; U14-01 aka T024
Non parametric filtering
Introduction to the SLAM problem
  • Friday May 28th, 2010; 09:30 - 12:30; Lab1401
laboratory (mlab, simple KF)
  • Friday June 4th, 2010; 09:30 - 12:30; Lab1401
laboratory (mlab, KF + more complex example, with EKF)
  • Monday June 7th, 2010; 09:30 - 12:30; Lab14a1
laboratory (mlab, conclusion example with EKF)

EKF vision-based SLAM

  • Wednesday June 9th, 2010; 09:30 - 12:30; U14-03 aka T014
Feature extraction
The Data Association Problem
Continuous Data Association
  • Wednesday June 9th, 2010; 14:30 - 17:30; U14-03 aka T014
The Loop Closing Problem
The Global Localization Problem
  • Thursday June 10th, 2010; 09:30 - 12:30; U14-01 aka T024
The SLAM scaling problem: Complexity and Consistency
Independent local maps
Conditionally independent maps
Large-Scale Visual SLAM
  • Thursday June 10th, 2010; 14:30 - 17:30; Lab1401
laboratory
  • Friday June 11th, 2010; 09:30 - 12:30; Lab14a1
laboratory
Motivazioni e obiettivi

Short description of the course

The course gives an introduction to Bayes Filtering, and presents an application of Extended Kalman Filtering to vision-based geometric modeling of a scene (a.k.a. SLAM), e.g., for mobile robotics applications. The EKF vision-based SLAM part will be given in consecutive days, so to allow students already skilled in Bayes Filtering to take only this part.

Registration

Please send an email to paola dot lembo at disco dot unimib dot it stating that you would like to participate to the course; please specify which part(s) of the course you are going to attend.

 

Grading

Grading will be based on the evaluation given by the lecturers on the quality of:

  • the documentation prepared by students on specific course topics;
  • the matlab programs that students will be asked to develop during the labs.

The grading scale adopted is:

A    90% and above (Excellence)     
B    80-89%            (Very Good)      
C    70-79%            (Good)           
D    60-69%             (Average)        
E    50-59%             (Unsatisfactory) 
F    49% and under  (Failure)        
Programma

Schedule

    Basics of probabilities
    Basics of filtering
    Gaussian Filtering (Kalman)
    Extended Kalman Filter
    Jacobian-based propagation of uncertainty
    EKF for parameter estimation with an implicit output transform
    Unscended Kalman Filter
    Information Form Gaussian Filtering
    Non parametric filtering
    Non parametric filtering
    Introduction to the SLAM problem
    Feature extraction
    The Data Association Problem
    Continuous Data Association
    The Loop Closing Problem
    The Global Localization Problem
    The SLAM scaling problem: Complexity and Consistency
    Independent local maps
    Conditionally independent maps
    Large-Scale Visual SLAM

 

Modalità di svolgimento

ECTS Notes

Here is our best estimate of the effort required to pass the course by the "average student". This estimate is provided for the students that need an estimate of the effort to get the credits.

  • Introduction to Bayes filtering
    1. the actual class hours: 12
    2. individual study for the class topics: 3
    3. actual lab. hours: 9
    4. individual working on the lab. assignments: 9
    5. total: 12+3+9+9 = 33
  • EKF vision-based SLAM
    1. actual class hours: 9
    2. individual study for the class topics: 3
    3. actual lab. hours: 6
    4. individual working on the lab. assignments: 9
    5. total: 9+3+6+9 = 27
Modalità d'esame
Materiale didattico

The course material, beside the references below, is available here; login and password have been communicated during classes. Notice though that most of the "Introduction to Bayes Filtering" part is perfectly covered by the first chapters of the book "Probabilistic Robotics".

  1. S. Thrun, W. Burgard, D. Fox. Probabilistic Robotics, MIT Press, September 2005.
  2. Conditionally Independent SLAM
  1. Inverse Depth Parametrization
Approfondimenti

Google Translate
Translate to English Translate to French Translate to German Translate to Spanish Translate to Chinese Translate to Portuguese Translate to Arabic
Translate to Albanian Translate to Bulgarian Translate to Croatian Translate to Czech Translate to Danish Translate to Dutch Translate to Finnish Translate to Greek Translate to Hindi
Translate to Hungarian Translate to Irish Translate to Japanese Translate to Korean Translate to Norwegian Translate to Polish Translate to Romanian Translate to Russian Translate to Serbian
Translate to Slovenian Translate to Swedish Translate to Thai Translate to Turkish

(C) Copyright 2016 - Dipartimento Informatica Sistemistica e Comunicazione - Viale Sarca, 336
20126 Milano - Edificio U14
redazioneweb@disco.unimib.it - ultimo aggiornamento di questa pagina 02/05/2012