The culture of coconut is economically relevant to Brazil but suffers with the breeding of pests that target coconut trees and bring on economic losses. One of the pests responsible for such is the palm weevil R. palmarum L., which is attracted to the infestation site by the odor released from the fermentation of plant tissues. This is particularly problematic when the crop, processing and disposal sites are in close location, for the residue discarded (i.e. green coconut shell) undergoes the action of microorganisms that produce volatile compounds and attract the pest to the crop site. This research had as its objective the isolation and characterization of bacteria and yeasts responsible for the fermentation of green coconut shell and production of volatile compounds that are attractive to the palm weevil. For such, green coconut shells were exposed to the local environment for the development of the fermentative processes, having the emitted volatiles analyzed on five different days of fermentation (0, 3, 5, 7, and 10 days). Thirty different compounds belonging to different chemical classes were detected for all of the analyzed days, being alcohols, ketones, aldehydes and hydrocarbons the most frequently found. Isolation procedures were performed immediately after the opening of the coconuts and at the sixth day of fermentation. After purification, thirty eight strains (15 bacteria and 23 yeasts) were obtained and characterized. The characterization of the strains was based on the determination of orphological, physiological and biochemical characteristics and the evaluation of the biotechnological potential. The isolated strains presented variable biochemical profile, although each of them displayed at least one positive result for the tested enzymes. Although bacterial isolates were halotolerant, yeast strains were more. Among the bacterial strains, the isolate A5 stood out for displaying activity for all of the tested enzymes except for laccase besides being capable of degrading filter paper in liquid media. The isolate E4 stood out among the yeast strains for presenting activity for five of the tested enzymes (amylases, pectinases, xylanases, proteases and lipases/esterases) besides displaying tolerance up to 10% NaCl. Therefore each strain presented different biotechnological potential, some of them being potential candidates for the production of volatile compounds by fermentation of the green coconut shell. A few strains were identified to the genus level by the technique of MILDI-TOF, resulting in identification of three bacterial (Bacillus, Pantoea e Arthrobacter) e three yeasts (Pichia, Cryptococcus e Rhodotorula) genera. The production of volatile compound by four selected strains was investigated and resulted on the detection of over 100 different compounds, among which are included ketones, carboxylic acids, pyrazines, esters, hydrocarbons, aldehydes, terpenes and terpenoids. A few of the detected volatiles had been previously reported as attractive to R. palmarum confirming that there is a relationship between the metabolic activity of the isolated microorganisms and production of volatiles attractive to this pest. / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / A cultura do coqueiro é economicamente importante para o Brasil, mas sofre com a proliferação de pragas que almejam os coqueiros e causam prejuízo econômico. Uma das pragas é a broca-do-olho-do-coqueiro, R. palmarum L., que é atraído para os locais de infestação pelo odor de fermentação de tecidos vegetais. Isto é particularmente problemático quando os locais de plantação, de processamento dos frutos e de descarte de resíduos são próximos, pois o resíduo gerado (casca do coco-verde, por exemplo) sofre ação de microrganismos que produzem compostos voláteis e atraem o inseto para a plantação. Esta pesquisa teve como objetivo o isolamento e a caracterização de bactérias e leveduras responsáveis pela fermentação da casca do coco-verde com concomitante produção de compostos voláteis atrativos para a broca-do-olho-do-coqueiro. Para tanto, cascas de coco-verde recém-abertas foram expostas ao ambiente para desenvolvimento de processo fermentativo, sendo a emissão de voláteis analisada em cinco dias diferentes de fermentação (dias 0, 3, 5, 7 e 10). Trinta compostos de classes químicas variadas foram detectados para os diferentes dias de fermentação, sendo encontrados álcoois, cetonas, aldeídos e hidrocarbonetos com maior frequência. Isolamentos foram realizados após abertura dos cocos e no sexto dia de fermentação. Após purificação, trinta e oito linhagens (15 bactérias e 23 leveduras) foram obtidas e caracterizadas. A caracterização teve como base a determinação de características morfológicas, fisiológicas e bioquímicas e a avaliação do potencial biotecnológico. As linhagens isoladas exibiram perfil bioquímico variado, sendo que todos apresentaram pelo menos um resultado positivo para alguma das enzimas extracelulares testadas. Embora as bactérias tenham sido halotolerantes, as leveduras foram mais. Dentre as bactérias isoladas, o isolado A5 se destacou por apresentar atividade para todas as enzimas testadas exceto lacases, além de demostrar capacidade de degradar papel de filtro em meio líquido. O isolado E4 se destacou entre as leveduras por apresentar atividade para cinco enzimas (amilases, pectinases, xilanases, proteases e lipases/esterases), além de apresentar tolerância a 10% de NaCl. Portanto, cada linhagem isolada apresentou potencial biotecnológico variável, algumas sendo potenciais candidatas para a produção de voláteis pela fermentação da casca do coco-verde. Alguns isolados foram identificados a nível de gênero pela técnica de MALDI-TOF, resultando na identificação de três gêneros de bactérias (Bacillus, Pantoea e Arthrobacter) e três de leveduras (Pichia, Cryptococcus e Rhodotorula). A produção de voláteis por culturas puras de quatro isolados selecionados foi estudada, sendo encontrados mais de 100 compostos diferentes, principalmente cetonas, ácidos carboxílicos, pirazinas, ésteres, hidrocarbonetos, aldeídos, terpenos e terpenóides. Alguns dos voláteis detectados já haviam sido relatados como atrativos para o R. palmarum., confirmando que existe relação entre a atividade metabólica dos microrganismos isolados e produção de voláteis atrativos para esta praga.

Submitted by JÚLIO HEBER SILVA (julioheber@yahoo.com.br) on 2017-05-02T17:14:06Z No. of bitstreams: 2 Dissertação - Josiane Moreira Cardoso - 2017.pdf: 2032261 bytes, checksum: 74966b401eaefa19e8806d5ea51fd25b (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2017-05-03T11:49:31Z (GMT) No. of bitstreams: 2 Dissertação - Josiane Moreira Cardoso - 2017.pdf: 2032261 bytes, checksum: 74966b401eaefa19e8806d5ea51fd25b (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2017-05-03T11:49:31Z (GMT). No. of bitstreams: 2 Dissertação - Josiane Moreira Cardoso - 2017.pdf: 2032261 bytes, checksum: 74966b401eaefa19e8806d5ea51fd25b (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2017-04-07 / Outro / In this work, experiments were performed to study the potential of coconut fiber (Cocos nucifera L.) as adsorbent for removing the pesticide parathion methyl from water. The analytical methodology used to quantify the pesticide in the adsorption studies was based on solid phase extraction (SPE) and gas chromatography with electron capture detection (GC/ECD). The method were validated and showed good analytical features in terms of selectivity, linearity, accuracy (recovery: 100.4%), precision (coefficient of variation: 9.63%;), limit of detection (0.003 mg L-1 ) and quantification (0.010 mg L-1 ). The matrix effect in pesticide quantification was evaluated and the results showed an enhanced chromatographic response for the parathion methyl in all the concentrations evaluated due to interference of endogenous coconut fiber compounds present in the solution after adsorption tests, therefore, the quantification was made using matrix-standard calibration solutions to compensate for matrix-induced effects and to obtain more accurate results. The adsorbent characterization (FTIR, MEV, surface area and pHpcz) showed that coconut fiber surface is composed of very irregular particles, with many cavities (pore diameter: 3.48 nm) and external surface area of 3.64 m2 g -1 . The presence of several functional groups: carboxylate, hydroxyl, carboxyl, among others was observed. The pH at point of zero charge (pHpcz) for the adsorbent was 5.85. The adsorption capacity of the coconut fiber was evaluated under different parameters: adsorbent treatment, adsorbent dosage, contact time and concentration of the pesticide. The adsorbent treatment with water, HCl or NaOH did not affect the adsorption capacity (75.6% of removal was obtained, independently of the treatment). The increasing in the adsorbent dosage of 5 to 20 g L-1 caused an increasing in the adsorption capacity from 25.0% to 76.6%. The adsorption equilibrium process was reached after 90 minutes of contact, with 86.1% of removal. The kinetic study showed that the Avrami kinetic model was the best fit for the experimental data. The adsorption capacity improved as function of the pesticide concentration increasing. The experimental data were fitted to Langmuir, Freundlich and Sips isotherm models and the best fit was obtained with Freundlich and Sips models. The maximum adsorption capacity (Qmax) were 39.8547 mg g-1 . Additional tests showed that the adsorption capacity was slightly affected when using real samples (natural water) in the adsorption experiments (82.24% removal) and presented excellent adsorption capacity when used in an alternative column adsorption procedure (85.73% removal). The results indicate that the coconut fiber has a great potential to be used as adsorbent material for the treatment of water contaminated with the pesticide methyl parathion. / Neste trabalho foi estudada a utilização da fibra de coco (mesocarpo do fruto de Cocos nucifera L.) como adsorvente para a remoção do agrotóxico parationa metílica de meio aquoso. A metodologia analítica empregada para quantificar o agrotóxico nos estudos de adsorção foi baseada nas técnicas de extração em fase sólida (SPE) e cromatografia gasosa com detecção por captura de elétrons (GC/ECD). O método foi validado e apresentou boa performance analítica em termos de seletividade, linearidade, exatidão (recuperação: 100,4%), precisão (coeficiente de variação: 9,63%), limites de detecção (0,003 mg L-1 ) e quantificação (0,010 mg L-1 ). Foi avaliado o efeito de matriz na quantificação do agrotóxico, e os resultados mostraram que ocorreu aumento na resposta cromatográfica para a parationa metílica em todas as concentração avaliadas devido à interferência de compostos endógenos da fibra de coco, presentes na solução após os ensaios de adsorção, desse modo, a quantificação foi realizada utilizando padrões analíticos preparados em extrato branco da matriz para compensar os efeitos de aumento na resposta cromatográfica induzido pela matriz e assim obter resultados quantitativos mais exatos. A caracterização do adsorvente (FTIR, MEV, área superficial e pHpcz) mostrou que a superfície da fibra de coco é constituída de partículas bastante irregulares e com muitas cavidades (diâmetro do poro: 3,48 nm), apresentando área superficial externa de 3,64 m2 g -1 . Foi observada a presença dos grupos funcionais: carboxilato, hidroxila, carboxila, entre outros. O pH no ponto de carga zero (pHpcz) para o adsorvente foi 5,85. A capacidade de adsorção da fibra de coco foi avaliada sob diferentes parâmetros: tratamento do adsorvente, dose de adsorvente, tempo de contato e concentração do agrotóxico. O tratamento do adsorvente com água, HCl ou NaOH não afetou a capacidade de adsorção, obtendo-se cerca de 75,6% de remoção, independentemente do tratamento. O aumento na dose de adsorvente de 5 para 20 g L-1 fez com que a capacidade de adsorção aumentasse de 25,0% para 76,6%. O equilíbrio do processo de adsorção foi alcançado após 90 minutos de contato, com 86,1% de remoção. O estudo cinético do processo mostrou que o modelo cinético de Avrami foi o que melhor ajustou aos dados experimentais. A capacidade de adsorção melhorou em função do aumento na concentração da parationa metílica no meio aquoso. Os dados experimentais foram ajustados às isotermas de Langmuir, Freundlich e Sips, sendo que os melhores ajustes foram observados para os modelos de Freundlich e Sips. A capacidade máxima de adsorção (Qmáx) foi 39,8547 mg g-1 . Ensaios adicionais mostraram que a capacidade de adsorção da fibra de coco foi pouco afetada quando foi utilizada amostra real (água natural) nos ensaios de adsorção (82,24% de remoção) e apresentou excelente capacidade de adsorção quando empregado em procedimento alternativo de adsorção em coluna (85,73% de remoção). Os resultados obtidos indicam que a fibra de coco pode ser utilizada como material adsorvente no tratamento de meio aquoso contaminado com o agrotóxico parationa metílica.

No presente trabalho, tanino (macromolécula de origem natural) foi utilizado como substituinte de fenol na formulação de matriz fenólica usada na preparação de compósitos, o que é possível devido à presença de anéis fenólicos sua estrutura. Os compósitos de matriz taninofenólica (50% em massa de tanino) apresentaram propriedades mecânicas superiores aquelas dos compósitos de matriz fenólica, quando reforçados pelo mesmo tipo de fibra, mostrando que a substituição de material obtido em larga escala a partir de fonte fóssil (fenol) é viável e pode ser feita por material obtido de fonte renovável (tanino) sem comprometimento de propriedades. Ainda, avaliou-se para propriedades de compósitos com diferentes tipos de reforços (fibras e cargas) através de caracterizações via análise térmica (TG e DSC), análise dinâmico-mecânica, espectroscopia na região do infravermelho, resistência ao impacto, absorção de água e microscopia eletrônica de varredura. Os ensaios de resistência ao impacto indicaram uma melhora de propriedades mecânicas quando da incorporação de fibras vegetais (juta e coco) nos termorrígidos fenólico e taninofenólico, além de mostrar que as fibras de juta reforçam as matrizes taninofenólicas mais eficientemente que as fibras de coco. As cascas da árvore de Acácia Negra, ricas em taninos, também foram utilizadas como agentes de reforço em compósitos na forma de fibras e cargas, obtidos através de desfibramento e pulverização, respectivamente. A utilização destes reforços, em diferentes proporções de massa, em compósitos de matriz taninofenólica (50% de tanino) levou a diferenças nas propriedades do compósito, com destaque para baixa absorção de água. Provavelmente, a presença de taninos no reforço e na matriz levou a intensas interações na interface fibra/matriz, diminuindo o número de vazios que podem alojar moléculas de água. As fibras de coco foram tratadas por ultra-som, a fim de avaliar a influência deste tratamento nas propriedades da fibra e dos compósitos reforçados por elas. As fibras de coco tratadas e não-tratadas foram caracterizadas quanto à composição química, índice de cristalinidade, ensaio de tração e microscopia eletrônica de varredura. Os resultados obtidos mostraram que o ultra-som é um tratamento promissor, pois modifica a morfologia da superfície das fibras sem levar à degradação química, sendo que a separação dos feixes de fibra que ocorre permite melhor interação fibra/matriz na interface, obtendo-se assim compósitos mais resistentes à água. Outros tratamentos utilizados em trabalhos anteriores a este, como mercerização, diminuem absorção de água dos compósitos por ela reforçados, ao custo de parcial degradação das fibras lignocelulósicas. No presente trabalho, materiais preparados a partir de matéria-prima oriunda de fontes renováveis foram processados, o que vem de encontro aos anseios atuais. Os compósitos obtidos têm potencial para aplicações não estruturais, por exemplo, em partes internas de veículos automotivos. / In the present work, tannin (macromolecule obtained from natural source) was used as substitute of phenol in the formulation of phenolic matrix composites, due to the presence of phenolic rings in its structure. The tanninphenolic matrix composites (50% w/w of tannin) presented mechanical properties better than those of phenolic matrix composites showing that substitution of material obtained in large scale from non-renewable source (phenol) can be done by material obtained from natural source (tannin) without compromising the properties of the composite. The tanninphenolic matrix composites reinforced by different reinforcing agents (fibers and particules) were characterized by different techniques: Izod impact strength, thermogravimetry (TG), differential scanning calorymetry (DSC), infrared spectroscopy (IV), dynamic-mechanical analysis (DMA) and scanning electron microscopy (SEM). The Izod impact strength showed an improvement of mechanical properties due to the incorporation of natural fibres (jute and coir) in the phenolic and tanninphenolic matrices and also the better reinforcement of these matrices by jute fibres, when compared to coir fibres. The barks of Acacia Mimosa (high content of tannin) were also used as reinforcing agents of the tanninphenolic matrices in the forms of fibres and particules. The presence of these reinforcing agents in the matrix led to differences in the properties of the composites, highlighted by its lower water uptake. The presence of tannins in both reinforcing agents and matrix enhanced the fiber/matrix interactions, lowering the voids that increase water uptake. The coir fibres were treated by ultrasound, in order to evaluate the influence of this treatment in the properties of the fibres and, therefore, the composites reinforced with them. Besides chemical composition, all the fibres were characterized by the following techniques: X-ray diffraction, tensile strengh, infrared spectroscopy (IV), and scanning electron microscopy (SEM). The results revealed that the ultrasound is a promising treatment of fibres for the processing of composites, because it modifies the morphology of the surface of fibres without leading them to chemical degradation. The separation of fiber beams allows enhancement of the fiber/matrix interactions, leading to composites with lower water absorption capacity. Other treatments, such as mercerization, for example, improved the impregnation of the fibres by the pre-polymer, leading to composites with better properties, at expenses of partial degradation of lignocellullosic fibres. In the present work, composites were prepared using material obtained from renewable source, according to the purpose of this work. The obtained composites presents potential for non-structural applications, such as, internal panels of cars, for example.

The consumption of trans fat is strongly associated to the development of cardiovascular disease (CVD) by its action on plasma lipids and because they act in an inflammatory processes; besides, vegetable oils are used by population to prevent these kind of disease. The objective of this research was to determine the effect of consuming a high fat diet supplemented with coconut oil, avocado oil and chia oil in inflammatory and metabolic parameters in wistar rats. After seven days in adjustment period, 35 male Wistar rats received a standard diet for 45 days (TC), hyperlipidemic diet and coconut oil (DHCO), hyperlipidemic diet and avocado oil (DHAB) and hyperlipidemic diet and chia oil (DHCH), by gavage in the amount of 0.8 mL/kg body weight. The diet consumption and serum TG levels were higher in TC. Liver weight was higher in groups DHCO and DHAB, with no statistical difference between groups (p <0.05) considering weight gain, epididymal fat weight and TC, HDL and LDL serum levels. Higher plasma concentrations of IL-1, IL-6, TNF-α and IFN-γ were found in the groups DHSF, DHCO and DHAB and lower in TC and DHCH groups. The IL-10 was higher in the group DHCH. In the morphometric analysis, the DHCH group showed increase on the cell number and reduction on their cytoplasm area. The intervention on the hyperlipidemic diet consumption with trans fatty acid allowed to observe that the lipid profile is preceded by a systemic inflammatory environment, which is modulated by the omega-3 fatty acid (α-linolenic acid), present in chia oil, which reversed this scene, but does not override the consumption of a balanced diet observed by the TC group. This proves that dietary intervention to prevent CVD must be associated to appropriate choices of lipid intake, maintaining an adequate intake of omega-3, as evidenced in this paper by the decrease of pro-inflammatory cytokines (IL-1, IL -6, TNF-α and IFN-γ) and IL-10 increased, which is considered anti-inflammatory. / O consumo de gordura trans é fortemente associado ao desenvolvimento de doenças cardiovasculares (DCV) por sua ação sobre os lipídios plasmáticos e por atuarem em processos inflamatórios, sendo que os óleos vegetais são utilizados pela população para a prevenção dessas doenças. O objetivo deste trabalho foi determinar o efeito do consumo de dieta hiperlipídica suplementada com óleo de coco, óleo de abacate e óleo de chia nos parâmetros metabólicos e inflamatórios em ratos wistar. Após sete dias em período de adaptação, 35 ratos wistar machos receberam por 45 dias dieta padrão (TC), dieta hiperlipídica e soro fisiológico (DHSF), dieta hiperlipídica e óleo de coco (DHCO), dieta hiperlipídica e óleo de abacate (DHAB) e dieta hiperlipídica e óleo de chia (DHCH), sendo que os óleos foram ofertados por gavagem na quantidade de 0,8 mL/kg de peso corporal. O consumo da dieta e níveis séricos de TG foram maiores no TC. O peso do fígado foi maior nos grupos DHCO e DHAB, sendo que não houve diferença estatística entre os grupos (p< 0,05) quanto ao ganho de peso, peso de gordura epididimal e níveis séricos de CT, HDL e LDL. Concentrações plasmáticas mais elevadas de IL-1, IL-6, TNF-α e IFN-γ foram encontradas nos grupos DHSF, DHCO e DHAB e as menores nos grupos TC e DHCH. A IL-10 foi maior no grupo DHCH. Na análise morfométrica, o grupo DHCH apresentou aumento do número de células e redução na área do citoplasma das mesmas. A intervenção do consumo de dieta hiperlipídica com ácido graxo trans permitiu observar que a alteração do perfil lipídico é precedida de um ambiente inflamatório sistêmico, o qual é modulado pelo ácido graxo ômega 3 (α-linolênico), presente no óleo de chia, o qual reverteu este quadro, mas não sobrepõe-se ao consumo de uma dieta equilibrada observada pelo grupo TC. Isto comprova que a intervenção dietética para prevenir a ocorrência de DCV deve estar associada a escolhas adequadas da ingestão de lipídios, mantendo um aporte adequado de ácido graxo ômega-3, evidenciado no trabalho pela diminuição das citocinas pró-inflamatórias (IL-1, IL-6, TNF-α e IFN-γ) e aumento da IL-10 considerada anti-inflamatória.

Indiana University-Purdue University Indianapolis (IUPUI) / An RGBZ synthetic dataset consisting of five object classes in a variety of virtual environments and orientations was combined with a small sample of real-world image data and used to train the Mask R-CNN (MR-CNN) architecture in a variety of configurations. When the MR-CNN architecture was initialized with MS COCO weights and the heads were trained with a mix of synthetic data and real world data, F1 scores improved in four of the five classes: The average maximum F1-score of all classes and all epochs for the networks trained with synthetic data is F1∗ = 0.91, compared to F1 = 0.89 for the networks trained exclusively with real data, and the standard deviation of the maximum mean F1-score for synthetically trained networks is σ∗ = 0.015, compared to σ_F1 = 0.020 for the networks trained exclusively with real F1 data. Various backgrounds in synthetic data were shown to have negligible impact on F1 scores, opening the door to abstract backgrounds and minimizing the need for intensive synthetic data fabrication. When the MR-CNN architecture was initialized with MS COCO weights and depth data was included in the training data, the net- work was shown to rely heavily on the initial convolutional input to feed features into the network, the image depth channel was shown to influence mask generation, and the image color channels were shown to influence object classification. A set of latent variables for a subset of the synthetic datatset was generated with a Variational Autoencoder then analyzed using Principle Component Analysis and Uniform Manifold Projection and Approximation (UMAP). The UMAP analysis showed no meaningful distinction between real-world and synthetic data, and a small bias towards clustering based on image background.

No description available.

[pt] Resíduos de biomassas são de grande interesse por serem matérias primas para a geração de energia renovável. Neste trabalho foram estudados os processos de conversão térmica de pirólise lenta e de carbonização hidrotermal, empregando os resíduos de bagaço de cana-de-açúcar, bagaço de malte e casca de coco. A cinética e a termodinâmica das reações foram determinadas experimentalmente pelo cálculo da energia de ativação, fator de frequência, energia livre de Gibbs, entalpia e entropia do complexo ativado. A pirólise foi avaliada empregando a termogravimetria em atmosfera inerte, para uma faixa de temperatura do ambiente até 1.000 graus Celsius e aplicando diferentes taxas de aquecimento. A carbonização hidrotermal procedeu em um reator tipo autoclave Parr modelo 452HC2, em diferentes temperaturas e tempos de operação. Modelos cinéticos da pirólise para métodos Model-free, denominados Kissinger, Flynn- WallOzawa (FWO) e Kissinger-Akahira-Sunose (KAS), e Model-fitting, intitulado Coats-Redfern, foram investigados. Para os modelos KAS, FWO e Coats-Redfern foram analisados 19 tipos de mecanismos reacionais e respectivos coeficientes de determinação (R2) foram determinados. O bagaço de cana-de-açúcar se ajustou ao método de Kissinger (R2 de 0,9973) e ao Coats-Redfern, entretanto os métodos FWO e KAS não se adequam a este material. Os testes com bagaço de malte e casca de coco se ajustaram a todos os métodos aplicados, apresentando valores de R2 elevados (0,9 a 0,9999). O modelo para a cinética da carbonização hidrotermal foi aplicado, onde as três biomassas apresentaram a ordem variando de 3 a 3,49, valor compatível com a literatura. Adicionalmente foram feitas caracterizações físico-químicas nos resíduos de biomassa e nos biocarvões produzidos na carbonização hidrotermal, incluindo análise elementar e imediata, microscopia eletrônica de varredura, espectroscopia de infravermelho, densidade aparente, pH, condutividade e poder calorífico. / [en] Biomass residues are of great interest because they are raw materials for the generation of renewable energy. In this work, the processes of thermal conversion of slow pyrolysis and hydrothermal carbonization were studied using the residues of sugarcane bagasse, malt bagasse and green coconut shell. The kinetics and thermodynamics of the reactions were determined experimentally by calculating the activation energy, frequency factor, Gibbs free energy, enthalpy and entropy of the activated complex. Pyrolysis was evaluated using thermogravimetry in an inert atmosphere, from a room temperature up to 1,000 Celsius and applying different heating rates. Hydrothermal carbonization was carried out in a Parr autoclave type 452HC2 reactor, at different temperatures and times of operation. Kinetic models of pyrolysis for Model-free methods, called Kissinger, Flynn-WallOzawa (FWO) and Kissinger-Akahira-Sunose (KAS), and Model-fitting, entitled Coats-Redfern, were investigated. For the KAS, FWO and Coats-Redfern models, 19 types of reaction mechanisms were analyzed and their determination coefficients (R2) were evaluated. Sugarcane bagasse was adjusted to the Kissinger method (R2 equal to 0.9973) and Coats-Redfern, however the FWO and KAS methods are not suitable for this material. The tests with malt bagasse and coconut fiber were adjusted to all applied methods, showing high R2 values (0.9 to 0.9999). The model for the kinetics of hydrothermal carbonization was applied, where the three biomasses presented the order varying from 3 to 3.49, a value compatible with the literature. In addition, physical-chemical characterizations were carried out for the biomass residues and biochar produced by hydrothermal carbonization, including elementary and immediate analysis, scanning electron microscopy, infrared spectroscopy, apparent density, pH, conductivity and calorific value.

The Simultaneous Localisation And Mapping (SLAM) problem is an open fundamental problem in autonomous mobile robotics. One of the latest most researched techniques used to enhance the SLAM methods is instance segmentation. In this thesis, we implement an instance segmentation system using Swin Transformer combined with two of the state of the art methods of instance segmentation namely Cascade Mask RCNN and Mask RCNN. Instance segmentation is a technique that simultaneously solves the problem of object detection and semantic segmentation. We show that depth information enhances the average precision (AP) by approximately 7%. We also show that the Swin Transformer backbone model can work well with depth images. Our results also show that Cascade Mask RCNN outperforms Mask RCNN. However, the results are to be considered due to the small size of the NYU-depth v2 dataset. Most of the instance segmentation researches use the COCO dataset which has a hundred times more images than the NYU-depth v2 dataset but it does not have the depth information of the image.

This dissertation draws upon Western literature in critical theory, aesthetics, art theory, and art history to explore how lying can foster aesthetic experience and the sociopolitical effects of this experience. It nominates the idea of pseudology—lying as an art—and outlines its distinguishing features from the dawn of postmodernism to contemporary practices. This study demonstrates an analysis of lying premised on an understanding of aesthetics as caught up in the wider issues of public pedagogy and everyday politics. Taking as case studies specific works of Marcel Duchamp, Robert Rauschenberg, VALIE EXPORT, and Carol Duncan, this dissertation argues for the narrative framing of artwork as paramount for its reception. As well, by examining the artistic mystifications of Mierle Laderman Ukeles, Guillermo Gómez-Peña, Coco Fusco, Joshua Schwebel, and Iris Häussler, this dissertation analyzes the use of pseudology in institutional critique. The study finds that perfidious practices can point to the importance of the relational boundary between what is real/unreal, highlight the social construction of this boundary’s aesthetic aspects, and reveal the ways in which each of us are active in the construction of a shared reality. Ultimately, our active framing of everyday life and the affective nature of our construction of a shared reality has been problematized by a contemporary prevalence of lying in the realms of public culture and politics. Pseudology reveals the power of narrative framing. The pseudological artworks discussed here expose, as models for the political aesthetic of lying, the need to debate the very tenets of reality constantly and continually—an essential civic action in the ethical, communal relationships of a democracy. / Thesis / Doctor of Philosophy (PhD) / An analysis of the use of lying as an artistic technique.

Instance segmentation has a great potential for improving the current state of littering by autonomously detecting and segmenting different categories of litter. With this information, litter could, for example, be geotagged to aid litter pickers or to give precise locational information to unmanned vehicles for autonomous litter collection. Land-based litter instance segmentation is a relatively unexplored field, and this study aims to give a comparison of the instance segmentation models Mask R-CNN and DetectoRS using the multiclass litter dataset called Trash Annotations in Context (TACO) in conjunction with the Common Objects in Context precision and recall scores. TACO is an imbalanced dataset, and therefore imbalanced data-handling is addressed, exercising a second-order relation iterative stratified split, and additionally oversampling when training Mask R-CNN. Mask R-CNN without oversampling resulted in a segmentation of 0.127 mAP, and with oversampling 0.163 mAP. DetectoRS achieved 0.167 segmentation mAP, and improves the segmentation mAP of small objects most noticeably, with a factor of at least 2, which is important within the litter domain since small objects such as cigarettes are overrepresented. In contrast, oversampling with Mask R-CNN does not seem to improve the general precision of small and medium objects, but only improves the detection of large objects. It is concluded that DetectoRS improves results compared to Mask R-CNN, as well does oversampling. However, using a dataset that cannot have an all-class representation for train, validation, and test splits, together with an iterative stratification that does not guarantee all-class representations, makes it hard for future works to do exact comparisons to this study. Results are therefore approximate considering using all categories since 12 categories are missing from the test set, where 4 of those were impossible to split into train, validation, and test set. Further image collection and annotation to mitigate the imbalance would most noticeably improve results since results depend on class-averaged values. Doing oversampling with DetectoRS would also help improve results. There is also the option to combine the two datasets TACO and MJU-Waste to enforce training of more categories.