humbleAR's profile - activity

Hi everyone,

I'm working on a project that involves several features related to face detection and some face information.

One of the features is detection of eyeglasses. This means that I have to make an algorithme that says YES or NO, the face is wearing glasses. I already have a face detector and eyes detectors that work well. We can see it as a detection problem or a classification problem.

I would like to have your opinion on what you consider as being a good method to solve that problem.

Than you much !

Hi everyone,

I'm working on a project that involves several features related to face detection and some face information.

One of the features is detection of eyeglasses. This means that I have to make an algorithme that says YES or NO, the face is wearing glasses. I already have a face detector and eyes detectors that work well. We can see it as a detection problem or a classification problem.

I would like to have your opinion on what you consider as being a good method to solve that problem.

Than you much !

Thank you for your answer Alexander. Getting deeper in OpenCV code, it seems that commenting the "clean" in the predict solves this specific problem. I will confirm after some in depth testing.

I have exactly the same problem but I don't see an answer for now.

Hello everyone,

I am working on an application on android that uses CvSVM class of OpenCV. I tested the code on my mac without any problem and it provides quite good results. The training and the detection part work quite fine, I tested them on mac OS and I included it in iOS without any problem.

I spent several weeks trying to include it in an android application using the NDK. Without Success. I first tried some basic opencv examples on android and it worked fine. But when using SVM, I have a SIGSEGV (Segmentation Fault) when it enters the predict (CvSVM::predict). There is a std::bad_alloc log.

Here is the most relevant log I got :

01-01 04:51:08.863: I/stderr(10641): terminate called after throwing an instance of 'std::bad_alloc'
01-01 04:51:08.863: I/stderr(10641):   what():  **std::bad_alloc**
01-01 04:51:08.867: A/libc(10641): **Fatal signal 11 (SIGSEGV)** at 0xdeadbaad (code=1), thread 10641 (mples.tutorial2)
01-01 04:51:09.004: I/DEBUG(91): *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** ***
01-01 04:51:09.582: I/DEBUG(91): becc40bc  53fb7451  /data/app-lib/org.opencv.samples.tutorial2-1/**libopencv_java.so (CvSVM::predict(float const*, int, bool) const)**

What I already tried (among other things):

• reducing image size to see if it is a memory problem
• verify all the classifiers and dictionary
• verify that there are actually detected key points

I really got our of ideas of what to try to make it work on android. I don't know if there is some restriction on android that makes it impossible to run. It works on iOS and on my Mac OS but not on android.

Thank you much for reading. Any help or suggestion is welcome.

Hello everyone,

I am currently working on a project that aims to detect fish and track them in a video. For example, draw a white box around a shark and red one around a Thuna. I have good quality pictures of the fish to track. And I'm trying to know which techniques may help me the best and provide robust tracking results on a video in acceptable response time (1 sec to detect is acceptable). There are some information that I think may be useful for you to help me :

• Colour should be taken in consideration because some fish species have same shape with different colors;
• Input video quality is quite good;

I would like to know which features would fit this application. I have thought for example to mix SURF and color image moments. Any ideas are welcome, if you can add examples of working similar application to your ideas, I would greatly appreciate it.

video frame :

image from database:

Thank you all.

Thanks a lot. Since some time passed, I have read some of the papers related to this project but I didn't read the last one that you give as reference. Thanks also for the very clear reconstruction of their approach. This is one of the most complete answers I have seen on a forum in my life.

Thanks a lot for the info, very helpful.

For example to detect planes that can be flying or landing or taking off...

Interesting, so, is there an alternative to that ? I mean, what to do to have something that is rotation invariant ? and as a bonus symmetry invariant ?

Hello everyone,

I understand that the HaarDetectObjects function based on Viola-Jones algorithm takes in consideration different window scales so I guess it is scale invariant. But Nothing being said about rotation invariance, I read that it's possible to make it rotation invariant by choosing the right Haar features (for example 4x rotated features by 45 degrees). But is it done by default in the provided XML file ?

Also, I would like to know, if it is symmetry invariant. For example, training on profiles of people looking to the right will work on people looking to the left ?

Thanks in advance.

OK @Rui Marques, I exchanged some mails with the guy who made the app on the video. He says that he simply used this implementation of KLT :

http://www.ces.clemson.edu/~stb/klt/

And just changed parameters and number of tracked points. Sounds almost too easy :D

You mean that he tracks fish using KLT but he doesn't recognize them ? That's not very useful in our case indeed. It's quite interesting to have your opinion on that because I don't and will not have hundreds of pics of every fish. But I think differences between fish species are bigger than differences between human (same specie) faces. You think it will require how many pics approximatively for every specie ? I wanted also to ask about optical flow, do you think it can be of any use here ? for tracking maybe ?

Anyway, I really like your explanations and advices @sammy.

Very useful comment @sammy , I guess it's better to use texture descriptors in our case, which one you think might be better here ?