Rank: 100 based on 539 downloads (last 30 days) and 4 files submitted
photo

Oswaldo Ludwig

E-mail
Company/University
ISR - University of Coimbra

Personal Profile:

Oswaldo Ludwig received the M.Sc. degree in electrical engineering from Federal University of Bahia, Brazil, in 2004, when he began to teach courses in electrical engineering, such as artificial intelligence, digital control, and digital signal processing. In 2012 he received the PhD degree in electrical engineering from University of Coimbra, Portugal, where he works as an invited assistant professor. His current research focuses on machine learning with application on several fields, such as pedestrian detection in the domain of intelligent vehicles.

Professional Interests:
Artificial Intelligence, Machine Learning

 

Watch this Author's files

 

Files Posted by Oswaldo View all
  • 1 - 4 of
  • 4
Updated   File Tags Downloads
(last 30 days) 		Comments Rating
11 Sep 2012 Feature selector based on genetic algorithms and information theory. The algorithm performs the combinatorial optimization by using Genetic Algorithms. Author: Oswaldo Ludwig information theory, genetic algorithms, mutual information, feature selection, artificial intelligen..., pattern recognition 86 6
  • 4.0
4.0 | 5 ratings
16 May 2012 Support Vector Neural Network (SVNN) A new training method for MLP neural networks based on the same principles of SVM. Author: Oswaldo Ludwig neural networks, machine learning, maximummargin princip..., svm, regularization, pattern recognition 63 4
17 Mar 2011 HOG descriptor for Matlab Image descriptor based on Histogram of Oriented Gradients for gray-level images Author: Oswaldo Ludwig image descriptor, artificial intelligen..., computer vision, i run your code, good 370 17
  • 4.18182
4.2 | 11 ratings
04 Nov 2010 MMGDX: a maximum-margin training method for neural networks Maximum-margin training method applicable to MLP in the context of binary classification. Author: Oswaldo Ludwig neural networks, machine intelligence, machine learning, artificial intelligen..., machine vision 20 1
Comments and Ratings by Oswaldo View all
  • 1 - 5 of
  • 9
Updated File Comments Rating
24 Jan 2013 HOG descriptor for Matlab Image descriptor based on Histogram of Oriented Gradients for gray-level images Author: Oswaldo Ludwig

Could you explain why it was adopted blockwise normalization? From the best of my knowledge, there is no mathematical foundation behind HOG descriptors, only good sense. This is my version, which was made available to be evaluated by the community. Go ahead, you should make the comparison with other algorithms in your case study (Science is not religion, feel free to doubt and inovate).

13 Dec 2012 Support Vector Neural Network (SVNN) A new training method for MLP neural networks based on the same principles of SVM. Author: Oswaldo Ludwig

Mahesh,

Regarding the application of my codes in multi-class problems, the feature selector based on genetic algorithms and information theory can deal with multi-class problem by adopting the target data, y, as a line vector in which each class is represented by a natural number, such as 1 for the first class, 2 for the second... The SVNN can be upgraded to solve multi-class problem by adapting the framework described in "K. Crammer and Y. Singer. On the Algorithmic Implementation of Multi-class SVMs, JMLR, 2001". However, since I finished my PhD, I have no time to work on this research line.
28 Sep 2012 HOG descriptor for Matlab Image descriptor based on Histogram of Oriented Gradients for gray-level images Author: Oswaldo Ludwig

Raul, the number of features depends on the number of cells and histograms bins, see Section II-A of this paper: http://www2.isr.uc.pt/~urbano/PDFs_PUB/Conferencias_Internacionais/2009/[113]2009ITSC09-129-FI.pdf
01 Jun 2012 HOG descriptor for Matlab Image descriptor based on Histogram of Oriented Gradients for gray-level images Author: Oswaldo Ludwig

An application for human detection (Windows only) can be found in http://webmail.isr.uc.pt/~oludwig/
The program is based on HOG descriptors for Matlab arranged in SRM-cascade, according to the methodology described in: O. Ludwig; “ Study on Non-parametric Methods for Fast Pattern Recognition with Emphasis on Neural Networks and Cascade Classifiers;” PhD Thesis, University of Coimbra, Coimbra, 2012. New images can be included in the folder "images" with the name "ped_XX". The algorithm was trained by using 18x36 gray-scale images from the Daimler Pedestrian Classification dataset. Double-click in detector_pkg.exe to install the program, then it is possible to run detector.exe.
16 May 2012 Support Vector Neural Network (SVNN) A new training method for MLP neural networks based on the same principles of SVM. Author: Oswaldo Ludwig

Dear Kamil,

Thank you for your interest in the code. Similarly the majority of the classifiers, the SVNN is problem dependent. MLPs trained by SVNN are likely to have better generalization capability than MLPs trained by usual gradient methods, such as GDX or LM, because SVNN is based on the maximum-margin principle. In relation to SVMs, SVNN can perform nonlinear classification faster than SVMs with nonlinear kernels, which have to perform a set of calculations per support vector, being especially hard when the number of support vectors is big (therefore, SVNN can be better for on-the-fly applications). Regarding the accuracy, both methods, SVM and SVNN, have good performance, which can vary according to the dataset. The SVNN training is quite similar to SVM training, due to a new regularization method named eigenvalue decay, which has an effect on the MLP classification margin similar to minimizing the norm of the parameters W, in the case of SVM.

Comments and Ratings on Oswaldo's Files View all
  • 1 - 5 of
  • 37
Updated File Comment by Comments Rating
14 Mar 2013 HOG descriptor for Matlab Image descriptor based on Histogram of Oriented Gradients for gray-level images Author: Oswaldo Ludwig tamanna

@Oswaldo Ludwig sir, may i have your full paper coding ''Trainable Classifier-Fusion Schemes: An Application To Pedestrian Detection''

24 Jan 2013 HOG descriptor for Matlab Image descriptor based on Histogram of Oriented Gradients for gray-level images Author: Oswaldo Ludwig Ludwig, Oswaldo

Could you explain why it was adopted blockwise normalization? From the best of my knowledge, there is no mathematical foundation behind HOG descriptors, only good sense. This is my version, which was made available to be evaluated by the community. Go ahead, you should make the comparison with other algorithms in your case study (Science is not religion, feel free to doubt and inovate).

24 Jan 2013 HOG descriptor for Matlab Image descriptor based on Histogram of Oriented Gradients for gray-level images Author: Oswaldo Ludwig Tencer, Lukas

Also you are doing normalization of the final vector, originally it supposed to be blockwise normalization, did you try that version? Why you decided for normalization only of final histogram.
24 Jan 2013 HOG descriptor for Matlab Image descriptor based on Histogram of Oriented Gradients for gray-level images Author: Oswaldo Ludwig Tencer, Lukas

Should not be contribution of individual points to the histogram weighted according to gaussian centred in the middle of each cell?
13 Dec 2012 Support Vector Neural Network (SVNN) A new training method for MLP neural networks based on the same principles of SVM. Author: Oswaldo Ludwig Ludwig, Oswaldo

Mahesh,

Regarding the application of my codes in multi-class problems, the feature selector based on genetic algorithms and information theory can deal with multi-class problem by adopting the target data, y, as a line vector in which each class is represented by a natural number, such as 1 for the first class, 2 for the second... The SVNN can be upgraded to solve multi-class problem by adapting the framework described in "K. Crammer and Y. Singer. On the Algorithmic Implementation of Multi-class SVMs, JMLR, 2001". However, since I finished my PhD, I have no time to work on this research line.
Top Tags Applied by Oswaldo
  • 1 - 5 of
  • 17
artificial intelligence, machine learning, neural networks, pattern recognition, computer vision
Files Tagged by Oswaldo View all
  • 1 - 4 of
  • 4
Updated   File Tags Downloads
(last 30 days) 		Comments Rating
11 Sep 2012 Feature selector based on genetic algorithms and information theory. The algorithm performs the combinatorial optimization by using Genetic Algorithms. Author: Oswaldo Ludwig information theory, genetic algorithms, mutual information, feature selection, artificial intelligen..., pattern recognition 86 6
  • 4.0
4.0 | 5 ratings
16 May 2012 Support Vector Neural Network (SVNN) A new training method for MLP neural networks based on the same principles of SVM. Author: Oswaldo Ludwig neural networks, machine learning, maximummargin princip..., svm, regularization, pattern recognition 63 4
17 Mar 2011 HOG descriptor for Matlab Image descriptor based on Histogram of Oriented Gradients for gray-level images Author: Oswaldo Ludwig image descriptor, artificial intelligen..., computer vision, i run your code, good 370 17
  • 4.18182
4.2 | 11 ratings
04 Nov 2010 MMGDX: a maximum-margin training method for neural networks Maximum-margin training method applicable to MLP in the context of binary classification. Author: Oswaldo Ludwig neural networks, machine intelligence, machine learning, artificial intelligen..., machine vision 20 1

Contact us