Submitted by Jos? Henrique Henrique (jose.neves@ufvjm.edu.br) on 2017-06-06T21:08:54Z No. of bitstreams: 2 jose_lucas_camargos.pdf: 1772896 bytes, checksum: ed61126d3964190579d1888b21edf3aa (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Rodrigo Martins Cruz (rodrigo.cruz@ufvjm.edu.br) on 2017-06-22T15:30:03Z (GMT) No. of bitstreams: 2 jose_lucas_camargos.pdf: 1772896 bytes, checksum: ed61126d3964190579d1888b21edf3aa (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2017-06-22T15:30:03Z (GMT). No. of bitstreams: 2 jose_lucas_camargos.pdf: 1772896 bytes, checksum: ed61126d3964190579d1888b21edf3aa (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2017 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior (CAPES) / Funda??o de Amparo ? Pesquisa do Estado de Minas Gerais (FAPEMIG) / Conselho Nacional de Desenvolvimento Cient?fico e Tecnol?gico (CNPq) / O objetivo deste trabalho foi avaliar a influ?ncia da redu??o de parcelas em algumas medi??es de invent?rio florestal cont?nuo na modelagem do crescimento e produ??o de um povoamento de Eucalyptus sp. O modelo de Clutter foi utilizado na modelagem do crescimento e da produ??o florestal, a partir de dados selecionados de 374 parcelas distribu?das em 172 talh?es. Foram utilizadas tr?s alternativas de modelagem: sem estratifica??o do povoamento, com estratifica??o do povoamento e um conjunto de 30 simula??es. Estas simula??es foram amostras de 40 parcelas aleat?rias cada, sem estratifica??o, com o objetivo de possibilitar uma an?lise geral dos efeitos do fen?meno estudado. Em todas as alternativas, foi realizada a modelagem do crescimento e da produ??o sem considerar redu??es nas medi??es de invent?rio florestal cont?nuo e considerando tr?s situa??es de redu??o. As situa??es consideradas foram medi??es de 25%, 50% e 75% das parcelas do invent?rio florestal cont?nuo nas medi??es dois e quatro, e 100% nas medi??es um, tr?s e cinco. Esta mesma metodologia foi repetida, por?m com a utiliza??o do modelo de regress?o n?o linear Log?stico, acrescido de um fator de corre??o. Ambos os modelos apresentaram boa precis?o para a predi??o e proje??o do volume do povoamento em idades futuras. O modelo Log?stico com fator de corre??o, entretanto, foi mais preciso que o modelo de Clutter. Para ambos os modelos foi constatado que n?o houveram perdas significativas de precis?o acarretadas pelas redu??es no n?mero de parcelas em algumas medi??es do invent?rio florestal cont?nuo. Para a modelagem do crescimento e da produ??o, ? recomendado que estas redu??es sejam criteriosas e que preferencialmente a ?ltima medi??o n?o seja reduzida. / Disserta??o (Mestrado) ? Programa de P?s-Gradua??o em Ci?ncia Florestal, Universidade Federal dos Vales do Jequitinhonha e Mucuri, 2017. / This study had as object to evaluate the influence of the reduction of plots at some measurements of continuous forest inventory in the growth and yield modeling of a Eucalyptus sp. forest stand. The Clutter model was used in forest growth and yield modeling, using selected data of 374 distributed plots in 172 forest compartments. It were used three modeling alternatives: without stratification of the forest stand, with stratification of the forest stand and a set of 30 simulations. These simulations were samples of 40 plots each one, without stratification, objectifying a general analysis about the studied phenomenon. To all the alternatives, it was made the forest growth and yield modeling without considering reductions in continuous forest inventory measurements and considering three reduction situations. The considered situations were measurements of 25%, 50% and 75% of the continuous forest inventory plots in the measures two and four, and 100% in the measures one, three and five. This same methodology was repeated, but using the Logistic nonlinear regression model, plus a correction factor. Both models had good precision to predicting and projecting the forest stand volume at future ages. The Logistic model with a correcting factor, however, had best precision than the Clutter model. For both models it was found that there was not significant losses of accuracy caused by reductions in the number of plots at some continuous forest inventory measurements. To forest growth and yield modeling, it is recommended for these reductions to be judicious and preferably the last measurement to not be reduced.

The focus of this thesis is on facial biometrics; specifically in the problems of face detection and face recognition. Despite intensive research over the last 20 years, the technology is not foolproof, which is why we do not see use of face recognition systems in critical sectors such as banking. In this thesis, we focus on three sub-problems in these two areas of research. Firstly, we propose methods to improve the speed-accuracy trade-off of the state-of-the-art face detector. Secondly, we consider a problem that is often ignored in the literature: to decrease the training time of the detectors. We propose two techniques to this end. Thirdly, we present a detailed large-scale study on self-updating face recognition systems in an attempt to answer if continuously changing facial appearance can be learnt automatically. / L'objectiu d'aquesta tesi és sobre biometria facial, específicament en els problemes de detecció de rostres i reconeixement facial. Malgrat la intensa recerca durant els últims 20 anys, la tecnologia no és infalible, de manera que no veiem l'ús dels sistemes de reconeixement de rostres en sectors crítics com la banca. En aquesta tesi, ens centrem en tres sub-problemes en aquestes dues àrees de recerca. En primer lloc, es proposa mètodes per millorar l'equilibri entre la precisió i la velocitat del detector de cares d'última generació. En segon lloc, considerem un problema que sovint s'ignora en la literatura: disminuir el temps de formació dels detectors. Es proposen dues tècniques per a aquest fi. En tercer lloc, es presenta un estudi detallat a gran escala sobre l'auto-actualització dels sistemes de reconeixement facial en un intent de respondre si el canvi constant de l'aparença facial es pot aprendre de forma automàtica.