Die Suche ergab 70 Treffer

12. Jul 2013 12:06
Forum: Computer Vision
Thema: Transform a4p3b
Antworten: 4
Zugriffe: 1148

Re: Transform a4p3b

Maybee I have todo some re-thinking (even so I really thought my soultion should be right xD) Thinking about the dimensions may help you. So in theory there should actually be two mappings, one for left and one for right. Why? Because you should also be able to create a panorama with different dime...
12. Jul 2013 11:02
Forum: Computer Vision
Thema: Transform a4p3b
Antworten: 4
Zugriffe: 1148

Re: Transform a4p3b

Hi, I try to transform the image a4p3b into the plane of a4p3a. Therefore, I proceed as stated in the task. First, I use the homography to get the coordinates of a4p3b in the plane of a4p3a. Afterwards, I use interp2 to interpolate the pixels of a4p3b at the calculated coordinates Looking at your r...
10. Jul 2013 19:57
Forum: Computer Vision
Thema: Robustness of RANSAC
Antworten: 1
Zugriffe: 742

Re: Robustness of RANSAC

OK, got it pretty robust, now. The trick was to choose a pretty conservative guess for the inlier probability (so that you have a bunch of iterations, though still below 100) together with a higher threshold for the chi-square-distances. You do not have to worry about a rather strict threshold, RANS...
8. Jul 2013 16:14
Forum: Computer Vision
Thema: Robustness of RANSAC
Antworten: 1
Zugriffe: 742

Robustness of RANSAC

Hi there, question: How exactly robust is RANSAC supposed to be? There is a couple of parameters for the stitching process, i.e. number of iterations (depending on our probability estimate of samples being inliers), the chosen threshold for small distances as well as the decision whether we accept a...
7. Jul 2013 13:05
Forum: Computer Vision
Thema: Typo in RANSAC Formular for Iterations
Antworten: 0
Zugriffe: 563

Typo in RANSAC Formular for Iterations

Howdy, just wanted to report a mistake in the slides: Page 26 of the l10-geometry -slides gives a formular to compute the number of iterations required to be (z*100)-% sure there is an uncontaminated sample. This formular should be k = \frac{\log(1 - z)}{\log(1 - w^k)} rather than k = \frac{\log(z)}...
1. Jul 2013 23:47
Forum: Computer Vision
Thema: A3 P2 - calculating class-conditional probabilities
Antworten: 2
Zugriffe: 979

Re: A3 P2 - calculating class-conditional probabilities

Hi, I just took a look at the lecture slides and, indeed, they are not quite clear about this. However, I think the lecture slides actually correspond to the assignment version just dropping the class' index. The major thing about Naive Bayes is its conditional independence assumptions.. and here we...
1. Jul 2013 18:08
Forum: Computer Vision
Thema: Assignment 3 - Problem 1B: What if C=0?
Antworten: 2
Zugriffe: 1004

Re: Assignment 3 - Problem 1B: What if C=0?

by definition C is the weight for \xi_i . Does this mean, if C=0 the result will be the same as in Problem1A? (...) Still, I'm getting some different values for w and b compared to 1A even if C is 0 You should not consider C to be a weight. It just describes the trade-off between the penalties of t...
28. Jun 2013 19:47
Forum: Computer Vision
Thema: Ex 3.2 - Comparing error rates
Antworten: 8
Zugriffe: 2054

Re: Ex 3.2 - Comparing error rates

As far as times are concerned, extracting the features takes most of the time (ca 1 minute for all images). KMeans takes about 4-6 seconds. All other computations are neglectable. Btw: My processor is a core 2 duo 2.5Ghz. I have nearly the same processor. Loading the images takes ca. 80 secs. and t...
28. Jun 2013 18:55
Forum: Computer Vision
Thema: compute histogram
Antworten: 2
Zugriffe: 841

Re: compute histogram

As is turns out there are two steps which involve computing a histogram. However, we are actually only concerned with one of them. 1) First, sift features are nothing else than histograms over the gradient discretized by a certain number of directions (this is already done by the given implementatio...
28. Jun 2013 18:40
Forum: Computer Vision
Thema: Assignment 3 Problem 1, warning using quadprog
Antworten: 3
Zugriffe: 1433

Re: Assignment 3 Problem 1, warning using quadprog

You can get rid of the warning by calling quadprog with:

Code: Alles auswählen

options = optimset('quadprog');
options.Algorithm = 'interior-point-convex';

28. Jun 2013 18:32
Forum: Computer Vision
Thema: Ex 3.2 - Comparing error rates
Antworten: 8
Zugriffe: 2054

Re: Ex 3.2 - Comparing error rates

Hi, thanks for that reply! I kind of have the same steps (even though I need some more lines ;) ). Just to clarify: You are working with feature vectors (128x1), right? So I have ~4800 feature vectors in total. For all these vectors I compute distances to the 50 means, search for the minimum, move ...
28. Jun 2013 18:12
Forum: Computer Vision
Thema: Ex 3.2 - Comparing error rates
Antworten: 8
Zugriffe: 2054

Re: Ex 3.2 - Comparing error rates

Hi, how do you solve k-means in 4-6 seconds?? I am working on it for days now, being really frustrated. It takes about 30-50 iterations (then it stops no matter what so it don't run forever). However, finding a cluster for each of the multiple thousand feature vectors (120 images * 10 and more vect...
28. Jun 2013 13:35
Forum: Computer Vision
Thema: Ex 3.2 - Comparing error rates
Antworten: 8
Zugriffe: 2054

Re: Ex 3.2 - Comparing error rates

Hi, I thought it would be interesting to compare the resulting error rates. What rates do you get and how do they vary (regarding different executions -> since other images are used for training)? I will post my rates soon (when I finished the task). Another question: Do your computations take some...
26. Jun 2013 16:01
Forum: Computer Vision
Thema: Assignment 3 2A - NaN values
Antworten: 2
Zugriffe: 650

Re: Assignment 3 2A - NaN values

Hi, my extract_feature class returns some training data that contains "NaN" values. What could cause this? I guess I am doing something wrong with applying SIFT, because my harris-feature points are correct (checked visually).I apply sift on the feature points. Is there anything I might have missed...
26. Jun 2013 14:50
Forum: Computer Vision
Thema: Ex 3.2 A - Boundary
Antworten: 1
Zugriffe: 557

Re: Ex 3.2 A - Boundary

Hi, bullet point two says: "...ignore the feature points within a 10-pixel wide boundary.".Does it address the image boundary? I just don't get what is meant. Can someone give me a hint :) Hi, you detect interest points by thresholding the Harris function values right? So this statement implies tha...