Projects

OverviewTopicsNotes & MaterialProjectsResources
Click for Course Overview
Click for Course Topics
Click for Course Notes& Material
Click for Course Projects
Click for Course Resources

Final project description can be found here

Projects 1-2-3:   

The projects are exercises from Rick Szeliski’s book: 
Computer Vision: 
Algorithms and Applications.

http://szeliski.org/Book/

Project 1:   Select one of the exercises from Section 2 (Image formation).
Submission date: March 25.

Project 2:   Select one of the following exercises from Section 4 (Feature detection and matching):
4.1, 4.2, 4.3+4.4 (solve 4.4 and evaluate the results using your solution of 4.3), 4.5, 4.8, 4.10,
Submission date: April 26.

Project 3:   Select one of the following exercises from Sections 6 or 8:
6.1, 6.2, 6.3, 6.4, 6.9, 8.2, 8.3, 8.4, 8.5
Submission date: May 24.

Instructions:

In each of the projects you are expected to do the following:

1. Solve the exercise

2. Implement your solution in Matlab. You will submit your code and data you used for testing/training your algorithm.

3. Submit a pdf files that includes
— 1-2 page summary of what you’ve done
— Results (images/graphs etc) that prove that you have succeeded in solving the exercise.
— What matlab version you have used
— What parts of your software were taken from others.

4. In the folder you submit there will be a file named runme.m. This is the file we will run to test your code.

Before you submit, please make sure that:
1. All the required files are included.
2. There are no “path” definitions that are unique to your own computer
3. Copy the submission folder onto a separate new directory and run it from there.

You will submit your solution (including the report) electronically. No need to print.
Best if you can email your projects to Ran Margolin<margolin@tx.technion.ac.il>.

 

Answers to some frequent questions: 

Q: What to bring?
A: I would recommend bringing a backup disk-on-key with the needed files in case the files you submitted prove insufficient.

Q: I chose a mostly analytical theoretical question with little Matlab. Is the demo still mandatory?
A: If your exercise has little Matlab and you feel that the demo would be redundant then let me know that you do not want to meet. Note that if you choose not meet, you still need to notify me by mail. 

Q:What does the demo consist?
A: The demo is a simple demonstration of your implementation. I might ask some clarifying questions to understand what you did or why, but nothing you need to prepare for…