Die Suche ergab 33 Treffer

von qgao
20. Jul 2011 12:46
Forum: Computer Vision
Thema: Assigment 4 - Problem 3
Antworten: 4
Zugriffe: 863

Re: Assigment 4 - Problem 3

tanne hat geschrieben:what is 'a small residual', up to which value is it correct?
I do not know how small a "correct" residual should be, as its value depends on a few factors (eg, distribution of 8 points, etc.).
von qgao
20. Jul 2011 12:39
Forum: Computer Vision
Thema: Assigment 4 - Problem 3
Antworten: 4
Zugriffe: 863

Re: Assigment 4 - Problem 3

The values that I got look like

1.0e-03 *

0.2332
0.9369
0.0898
0.1669
0.4917
0.1868
0.0003
0.0397

But NO guarantee of the correctness. -_-b
von qgao
20. Jul 2011 12:07
Forum: Computer Vision
Thema: A4A1 - harris parameters
Antworten: 4
Zugriffe: 664

Re: A4A1 - harris parameters

robert.n hat geschrieben:What are the definitions of testing and training error?
training error: rate of correct recognition w.r.t. training data;
testing error: rate of correct recognition w.r.t. testing data.
von qgao
20. Jul 2011 12:00
Forum: Computer Vision
Thema: A4 - Problem 3, Task B
Antworten: 2
Zugriffe: 598

Re: A4 - Problem 3, Task B

"Use the scheme from the 3rd assignment to nd the putative matches. Only consider the Harris points." What does that mean? Not to use the Harris-Laplace approach, but instead find corner points using harris IP detection (with only one scale)? // EDIT: I'm currently using harris-laplace. Do NOT use ...
von qgao
14. Jul 2011 17:27
Forum: Computer Vision
Thema: Problem with given sift.m
Antworten: 2
Zugriffe: 458

Re: Problem with given sift.m

tanne hat geschrieben:just drop all points beyond a certain treshold at the borders of the image, just like in the last exercise.
Yes, exactly.
von qgao
14. Jul 2011 17:26
Forum: Computer Vision
Thema: Image Derivate for Time..
Antworten: 2
Zugriffe: 502

Re: Image Derivate for Time..

tanne hat geschrieben:how do i calculate the derivate for the time from an image? i.e. the I_t of lucas-kanade.. i'm kinda stuck. :/
\(I_t \approx I(t+1) - I(t)\)
von qgao
16. Jun 2011 11:54
Forum: Computer Vision
Thema: chi²-Distance
Antworten: 1
Zugriffe: 443

Re: chi²-Distance

robert.n hat geschrieben:How is the definition of the chi²-distance?
You can find it in the slides of the 5th lecture.
von qgao
16. Jun 2011 11:51
Forum: Computer Vision
Thema: solution?
Antworten: 1
Zugriffe: 379

Re: solution?

tanne hat geschrieben:will the solution you hand out will be online aswell?
No, you can only get the printed version. Sorry.
von qgao
7. Jun 2011 19:13
Forum: Computer Vision
Thema: Exercise 3, Problem 1
Antworten: 8
Zugriffe: 1133

Re: Exercise 3, Problem 1

are the example pictures created by the same parameters as we should use, i.e. should our solution look exactly the same when we do it right? The example figures are the results of my quick implementation, but I am not sure they are correct. These figures are just used to let you know how to mark t...
von qgao
31. Mai 2011 12:44
Forum: Computer Vision
Thema: Exercise 2, Problem 4
Antworten: 3
Zugriffe: 592

Re: Exercise 2, Problem 4

Hi, The covariance matrix that we are calculating is 8064x8064. I am running into memory problem if I try to do an SVD on the matrix. Are we supposed to work with the covariance matrix for this problem the image matrix itself as with the last problem? Thanks You do not have to compute the whole cov...
von qgao
30. Mai 2011 11:19
Forum: Computer Vision
Thema: Exercise 2, Problem 4
Antworten: 3
Zugriffe: 592

Re: Exercise 2, Problem 4

Hello, I just understand the full task, that we should generate six random samples for six faces from our given dirstribution and we should consider 20 bases, alright? So we have to draw for each sample from 20 (and not 6) directions or have i understood the task wrong? Sorry, this is a typo error....
von qgao
25. Mai 2011 10:59
Forum: Computer Vision
Thema: Exercise 2, Problem 3
Antworten: 7
Zugriffe: 894

Re: Exercise 2, Problem 3

another question, Problem 4. for calculating the likelihood as it is described in the task, we now need the 'real' covariance matrix, 8064x8064. and we need the eigenvalues and eigenfaces of it, but the matrix is too big to use svd. and also we cant use our 20 dimension subspace of problem3, which ...
von qgao
24. Mai 2011 12:23
Forum: Computer Vision
Thema: Exercise 2, Problem 3
Antworten: 7
Zugriffe: 894

Re: Exercise 2, Problem 3

I think your understanding is right. :)
von qgao
18. Mai 2011 17:59
Forum: Computer Vision
Thema: Exercise 2, Problem 1
Antworten: 1
Zugriffe: 563

Re: Exercise 2, Problem 1

That sentence is really somewhat ambiguous.

Well, in other words, it is up to you whether you use for-loop or not.
von qgao
10. Mai 2011 19:16
Forum: Computer Vision
Thema: lecture/exercise tomorrow?
Antworten: 1
Zugriffe: 459

Re: lecture/exercise tomorrow?

Right. No lecture, no exercise class tomorrow.

But tomorrow there will be the exercise interviews in our Linux lab (Room 346, S3|05).

You can register here: http://goo.gl/MJEyv

Zur erweiterten Suche