Ask Your Question

Shin's profile - activity

2018-10-29 06:17:01 -0500 received badge  Nice Question (source)
2017-06-06 12:07:45 -0500 received badge  Notable Question (source)
2015-04-12 04:37:08 -0500 received badge  Popular Question (source)
2014-12-09 13:48:23 -0500 marked best answer Integrating different cues in a classification task

Is it possible to combine different features in a classification task? I read about training SVMs for each kind of collected descriptors and then use a weighted summation of the results to actually integrate/combine information, but how it should be done in practice?.

2014-12-09 13:45:14 -0500 marked best answer Opencv 2.4 : which language should i choose?

I know this question has been asked before, but i was wondering if with the current version of OpenCV (release 2.4) has changed anything.

Should i prefer the C++ or Python interface? Are there any big differences between them? Can you easily translate opencv code from one language to the other? And finally, can i use this book (amazon->Learning OpenCV) to get started anyway or are there any better ones?

Thank you.

2013-05-05 17:18:10 -0500 commented answer Precision-recall evaluation

I think that's it. Thank you for everything :-).

2013-05-05 16:58:37 -0500 commented answer Precision-recall evaluation

@unxnut: So, if for example, i'm using SURF features, i could try varying the hessian threshold?

2013-05-05 15:07:16 -0500 commented answer Precision-recall evaluation

Number of clusters (first step is a Bag-of-words classification approach)

2013-05-05 14:05:23 -0500 commented answer Precision-recall evaluation

For the "apple" classifier (a 1-vs-all classifier) i take precision and recall for different k values. I then plot them like in

2013-05-05 13:47:32 -0500 commented answer Precision-recall evaluation

I'm actually a little bit confused. Could you elaborate more or make an example? If i'm following your reasoning, you mean Recall = # True Positives / # Total Positives, but total should indeed be equal to (#True Positives + #False Negatives), which corresponds to my definition from earlier.

2013-05-05 13:07:20 -0500 commented answer Precision-recall evaluation

@unxnut: Precision = TP/(TP+FP), Recall = TP/(TP+FN)

2013-05-05 12:12:33 -0500 commented answer Precision-recall evaluation

@unxnut Hi unxnut. Thank you for your comment. I was wondering if you could take a quick look to my confusion matrixes. You can find a screenshot of my spreadsheet here

2013-05-04 11:39:27 -0500 asked a question Precision-recall evaluation

Is this

<a href=">

a possibile outcome for a precision-recall graph or you think i'm doing something wrong? (The blue line is relative to a one-step classifier, the other line refers to the same classifier with the difference that this time a second step is added in the classification process).

2013-03-26 17:05:58 -0500 asked a question From confusion matrix to ROC graph

Hello there, i recently implemented a Bag of Words categorization algorithm based on the one described in this paper

All works well, but i'd like to measure accuracies of the classifiers using ROC curves or, perhaps, precision-recall graphs.

I can easily get the confusion matrix for each of the classifiers but i don't know what parameter i should change to get more points and actually plot the curves.

Could someone please explain this to me?

2013-03-15 11:50:18 -0500 commented answer Global image feature implementation

@Guanta: thank you. I have seen that example, but i am interested in mixing different kinds of features, combining descriptors from, say, sift AND surf, in single vectors and then getting bow descriptors. Is this approach feasible or i'm saying something wrong ?

2013-03-15 07:10:35 -0500 commented answer Global image feature implementation

@Guanta: Thank you for your answer. Actually the BoW approach using local features was my first thought but I haven't found examples on how to use BoW with multiple kind of features. I'm sorry to ask but could you provide minimal sample code?

2013-03-14 16:33:16 -0500 commented answer Global image feature implementation

@StevenPuttemans : Thank you for your answer. I'd like to use a global descriptor for a color image in combination with a global descriptor for a depth image (from which i can get a point cloud) and combine them in a single descriptor vector. My ultimate goal is to feed the descriptor vectors to a support vector machine. For this task, i want the descriptor to be scale and rotation invariant. Any suggestion?

2013-03-14 14:15:51 -0500 commented answer Correct way to read depth images

Thank you. Well, i was reading them using a cv::Mat object of type CV_16UC1. I'd like to know if some scaling operation is needed (considering that these are indeed kinect depth images) and that most opencv functions work with CV_8U matrixes.

2013-03-14 14:14:29 -0500 asked a question Global image feature implementation

Is there any OpenCV implementation of a global image features detector?

2013-03-14 11:52:17 -0500 asked a question Correct way to read depth images

I need to use the dataset at and i'd like to know how to correctly read in the depth images using opencv.

I've uploaded one of these depth images here:

Thank you.