Curriculum Vitae (Lattes Plataform)
Current Research Projects
Dimensionality reduction using pixel clustering and principal components for pattern recognition.
(From 2015 until today)
Description:Pattern classification is the task that binds an object (sound, text, image or description) to a class from a pre-established set. It is an essential task in the Machine Learning field. Classification algorithms build automatic classification models from a tagged dataset. Each item in this set is a pair: object and class label. Some application for the classification task include: speaker recognition (identify the person that have these voice characteristics); sentiment analysis/opinion mining (classify the whether the opinion in the text is positive or negative); face verification (check if a face image of a person belongs to the person it claims to be); loan approval (decide whether to grant credit). Usually, a feature vector represents an object from the dataset. Each feature is a dimension of the problem. High dimensionality is a limiting factor for the classification task. It happens when the number of features for each example is so high that it becomes infeasible to execute the learning algorithm. In such case it is necessary to perform dimensionality reduction: find a new representation for an object with few features retaining relevant information for classification. In this project, we mainly deal with three methodologies for dimensionality reduction: Pixel Clustering, Fractional Eigenfaces, and Minimum Classification Error Principal Component Analysis. We want to use these methodologies to propose new dimensionality reduction techniques and machine learning algorithms such as deep neural networks. We will evaluate the proposed methods in well knows machine learning dataset and investigate new application using them.
Main publications
(Direct acces clicking the title)
DE CARVALHO, TIAGO B.A.; SIBALDO, MARIA A.A.; TSANG, ING REN; CAVALCANTI, GEORGE D.C.; SIJBERS, JAN; TSANG, ING JYH. IntensityPatches and RegionPatches for Image Recognition. APPLIED SOFT COMPUTING, v. 62, p. 176-186, 2018.
DE CARVALHO, TIAGO B. A.; SIBALDO, M. A. A.; TSANG, I. R.; CAVALCANTI, G. D. C. Principal Component Analysis for Supervised Leaning: a Minimum Classification Error Approach. Journal of Information and Data Management - JIDM, v. 8, p. 131-145, 2017. (PDF download)
PAES FERREIRA, FABRICIO; DE CARVALHO, TIAGO BUARQUE ASSUNÇÃO. Analysis of Wavelet Families for Face Recognition. In: 2017 Workshop of Computer Vision (WVC), 2017, Natal. 2017 Workshop of Computer Vision (WVC), 2017. p. 90-95.
NUNES, J. A. C.; FERREIRA, F. P.; DE CARVALHO, TIAGO B. A. Waveletfaces and Linear Regression Classification for Face Recognition. In: XIII Workshop de Visão Computacional - WVC 2017, 2017, Natal-RN. XIII Workshop de Visão Computacional - WVC 2017, 2017. p. 144-149.
DE CARVALHO, T. B. A.; SIBALDO, MARIA A. A.; TSANG, I.R.; CAVALCANTI, G. D. C.; TSANG, I.J.; SIJBERS, J. Pixel Clustering for Face Recognition. In: Brazilian Conference on Intelligent Systems (BRACIS), 2016, Recife. 2016 5th Brazilian Conference on Intelligent Systems (BRACIS), 2016. p. 121-126.
DE CARVALHO, T. B. A.; SIBALDO, MARIA A. A.; TSANG, I.R.; CAVALCANTI, G. D. C. Minimum Classification Error Principal Component Analysis. In: Symposium on Knowledge Discovery, Mining, and Learning (KDMile), 2016, Recife. PROCEEDINGS OF THE 4TH SYMPOSIUM ON KNOWLEDGE DISCOVERY, MINING, AND LEARNING, 2016. p. 68-75. (PFD download)
DE CARVALHO, T. B. A.; COSTA, A. M.; SIBALDO, M. A. A.; TSANG, I. R.; CAVALCANTI, G. D. C. Supervised fractional eigenfaces. In: 2015 IEEE International Conference on Image Processing (ICIP), 2015, Quebec City. 2015 IEEE International Conference on Image Processing (ICIP), 2015. p. 552-555.
SIBALDO, MARIA A. A.; DE CARVALHO, TIAGO B. A.; REN, TSANG ING; CAVALCANTI, GEORGE. Recommender System Based on Modularity. In: the 2014 Recommender Systems Challenge, 2014, Foster City. Proceedings of the 2014 Recommender Systems Challenge on - RecSysChallenge '14, 2014. p. 58.
LEITE, L. S. ; SIBALDO, M.A.A. ; DE CARVALHO, T.B.A. ; SOUZA, R. N. P. M. . Montanha de Chomsky: jogo tutor para auxílio no ensino de Teoria da Computação. In: Congresso da Sociedade Brasileira de Computação (Workshop sobre Educação em Computação), 2014, Brasília. XXXIV Congresso da Sociedade Brasileira de Computação. Porto Alegre: Sociedade Brasileira de Computaçã, 2014. p. 1364-1373.
DE CARVALHO, T. B. A.; SIBALDO, M. A. A.; TSANG, I. R.; CAVALCANTI, G. D. C.; TSANG, I. J.; SIJBERS, J. Fractional Eigenfaces. In: 2014 IEEE International Conference on Image Processing (ICIP), 2014, Paris. 2014 IEEE International Conference on Image Processing (ICIP). p. 258.
DE CARVALHO, T. B. A.; SIBALDO, M.A.A.; TENORIO, D.J.; TSANG, I.R.; CAVALCANTI, G.D.C.; TSANG, I.J. Neighborhood coding for image representation and neighborhood operations, 2012, Seoul, 2012.
DE CARVALHO, T. B. A.; TENORIO, D. J.; TSANG, I. R.; CAVALCANTI, G. D. C.; TSANG, I. J. Neighborhood coding for bilevel image compression and shape recognition. In: IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), 2010, Dallas, Texas, U.S.A. Acoustics Speech and Signal Processing (ICASSP), 2010 IEEE International Conference on, 2010. p. 1302-1305.