XE33IRO - Intelligent Robotics (winter 2004)1

Lecturer: Tomáš Pajdla  
Assistent: H. Bakstein  
Department: K333  

Lecture: Mo 11:00 - 12:30 in K108  
Labs: Po 18:00 - 19:30 in K132  

Up to date information at: http://cyber.felk.cvut.cz/teaching/.

The goal of the course

The goal of the course is to show how to solve an engineering problem comprising computer vision, robotics, and artificial intellignece and how to present the solution.


  Date Lecture Labs
1. 4.10. Introduction Test A
2. 11.10. Vision I. Consultation
3. 18.10. Kinematikcs I. Consultation
4. 25.10. Planning I. Consultation
5. 1.11. Vision II. Consultation
  4.11. 1. report submission deadline on 4.11. 12:00
6. 8.11. Planning II. Consultation
7. 15.11. Planning III. Consultation
8. 22.11. Vision III. Consultation
9. 29.11. Kinematics II. TEST B
  2.12. 2. report submission deadline on 2.12. 12:00
11. 6.12. Planning IV. Consultation
12. 13.12. Kinematics III. Consultation
13. 3.1 ``to be specified'' Consultation
14. 10.1 TEST C D E M O
  10.1 DEMO presentation and testing
14. 13.1.2005 3. (final) report submisison deadline 13.1.2004 12:00

Reports received in time will be evaluated according the eveluation scheme. For the final deadline, only reports received before deadline will contribute to the final grade.


Groups of three sutdents solve a project. To finish the semestra and to be allowed to come to the examination, students have to
  1. nad in project reports and
  2. demonstrate a program (DEMO) on test data.

Students will have allocated computersw during the labs. The lecturer and the teaching assistants will be available after making previously an appointment by email. Only formally correct parts of the reports will be discussed.

Groups of three students wil receive a group number. after sending their names by email to iro2004@cmp.felk.cvut.cz.

Project definition

1-st Lab:

Matlab and Robotic toolbox will be available at lab 132. To simulate robot movement in a constrained enviromant, functions simulating a proximity sensor and colisions with obstackles will be provided.

Project presentation

Project results will be presented as a research report. To be acepted, the report needs to be formally as well as technically correct

A technically correct report

  1. contains no errors
  2. proves or experimantally demonstrates statements, and
  3. contains sufficient information for others to repeat the work.

A formally correct report follows the rules at http://www.writing.eng.vt.edu/handbook/

Structure of the report.

Submitting the projects

Project reports must be submitted by in PDF format (from MS WOrd print to PDF in lab 132) by emailing it to iro2004@cmp.felk.cvut.cz before 13.1.2005 do 12:00. The email must comply the following rules (?? stands for the group number):

Submission not complying to the above rules will not be accepted.

DEMO will be demonstrated and tested during a session woth a teaching assistent after making an appointment with him any time before the deadline on 12.1.2005. Test data will be used to test the program in an unknown environment. Students may be asked to modify their program to be able to solve the tproblem in a new situation.

Projects: Projects and results.

Evaluation of reports

Formal and technical correctness of the reports are assesed independently in the scale 0-100 points. The resulting score is obtained as

\mbox{report score} = 0.5 (\mbox{technical scor...
...,(\mbox{formal score} \frac{\mbox{technical score}}{100}) .

The formal score is weighted by the technical score. If the report is formally incorrect, it still may receive apositive score fot its technical value provided that the length of the report is not exceeded.

See the following for scoring criteria scoring form (pdf).


An example of a test also in (pdf). Test results.

T E S T will be held during the lecture on 10.1.2005.


The final grade will be composed in the following way:

\mbox{grade} = 0.5 (\mbox{grade for report}) + 
0.15 (...
...grade for test C}) + 
0.2 (\mbox{grade for oral exam}) ,

and the mark will be computed as follows:
grade mark
90 - 100 1
70 - 89 2
50 - 69 3
00 - 49 unsatisfactory



R. Hartley and A. Zisserman.
Multiple View Geometry in Computer Vision.
Cambridge University Press, 2000.
(projective plane pp. 3-16, camera pp. 144-155, planar homography pp.184-185, 71-73, SVD 556-567).

Introduction to Matlab Example.

Perspective camera model.

On-line Robot Kinematics in Matlab Robot Toolbox.

Robot kinematics, Nov 1, 2004 lecture S1, S2, S3, S4.

State space search, Nov 8, 2004 lecture S5, S6, S7, S8, S9, S10,

MotionLPanning Recources


... 1
PDF also available

Tomas Pajdla 2004-12-06