Browsing by Issue Date, starting with "2024-01"
Now showing 1 - 10 of 33
Results Per Page
Sort Options
- «O tempo de falar chegou!» Significado e importância teológico-política do Apelo à nobreza cristã da nação alemã (12 de agosto de 1520)Publication . Rosa, José Maria SilvaO texto Apelo à nobreza cristã da nação alemã, de agosto de 1520, é um dos mais importantes escritos reformadores de Martinho Lutero (juntamente com Da catividade babilónica da Igreja, de outubro de 1520 e Acerca da liberdade do cristão, de novembro de 1520, entre outros). Texto de teor panfletário, não obstante as questões prementes do tempo a que queria responder, representa também um ponto de chegada nas disputas sobre as relações entre os poderes temporal e espiritual (De potestate), que tinham alimentado o debate teológico-político na Baixa Idade Média, nomeadamente no que se refere ao primado do concílio sobre o Papa (conciliarismo), primazia reconhecidamente antiga, e que, pelo menos desde João Quidort de Paris (1273-1306) a João de Gershom (1362--1429), contrariava a reivindicação papal da plenitude do poder (plenitudo potestatis papalis), não só na esfera dos assuntos temporais terrenos, mas também no plano interno da organização e administração da Igreja. A leitura papalista transgredia uma ideia do Papa Gelásio, no final século v, sobre a separação dos dois poderes ou dois gládios: o temporal e o espiritual, conceção sucessivamente traída ao longo de séculos pelo chamado augustinismo político, o qual, muito significativamente, procurava apoio no livro XIX de A cidade de Deus de Santo Agostinho. Não obstante, Lutero é também um leitor (parcial) de Santo Agostinho e de São Paulo e, como tal, procura pensar a Igreja não a partir da separação, oposição ou subordinação dos poderes, mas da graça e dos carismas (entre os quais o do dominium temporal de um dominus num reino), os quais são indispensáveis para a edificação do único Corpo de Cristo, que é simultaneamente temporal e eterno. Por este viés, todos os ministérios, inclusive aqueles considerados mais humildes, ganham grande valor espiritual (ou secularizam-se, noutra perspetiva). Seja como for, na sequência da Querela das Indulgências, o apelo de Lutero que incita os príncipes alemães à autonomização face ao poder secular de Roma dá espaço depois (à custa de muito sonho e de muito sangue, tem de se dizer) à afirmação dos Estados-nação do sacro-império romano-germânico. Ao mesmo tempo, por via do primado do «sacerdócio comum dos fiéis» frente ao poder magisterial do «sacerdócio ordenado» (potestas ordinis), há neste opúsculo, no que se refere à interpretação e pregação das Escrituras, um potencial muito virulento, explosivo, e mesmo apocalíptico (estamos no fim dos tempos, o Anticristo está à porta, etc.), que, se politicamente rebenta logo nos anos seguintes (1524-1525), nos campos da Alemanha, já de um ponto de vista hermenêutico mais amplo, recupera alguns processos de subjetivação e de apropriação dos «lugares de fala», bem como conceções nominalistas que já vinham ganhando relevância nos séculos xiv e xv, procedimentos que irão aprofundar-se quer na modernidade racionalista (do cogito cartesiano ao Ich denken kantiano) quer na modernidade fideísta do credo (Jansénio, Pascal, etc.). Não obstante, o que se impõe de forma premente, em 1520, é reformar a Igreja, pois «acabou o tempo do silêncio e chegou o tempo de falar!» («Die Zeit zu reden ist kommen»). E é esta decisão de Lutero de tomar palavra o que mais nos importa.
- 6D Pose Estimation and Object RecognitionPublication . Pereira, Nuno José Matos; Alexandre, Luís Filipe Barbosa de Almeida6D pose estimation is a computer vision task where the objective is to estimate the 3 degrees of freedom of the object’s position (translation vector) and the other 3 degrees of freedom for the object’s orientation (rotation matrix). 6D pose estimation is a hard problem to tackle due to the possible scene cluttering, illumination variability, object truncations, and different shapes, sizes, textures, and similarities between objects. However, 6D pose estimation methods are used in multiple contexts like augmented reality, for example, where badly placed objects into the real-world can break the experience of augmented reality. Another application example is the use of augmented reality in the industry to train new and competent workers where virtual objects need to be placed in the correct positions to look like real objects or simulate their placement in the correct positions. In the context of Industry 4.0, robotic systems require adaptation to handle unconstrained pick-and-place tasks, human-robot interaction and collaboration, and autonomous robot movement. These environments and tasks are dependent on methods that perform object detection, object localization, object segmentation, and object pose estimation. To have accurate robotic manipulation, unconstrained pick-and-place, and scene understanding, accurate object detection and 6D pose estimation methods are needed. This thesis presents methods that were developed to tackle the 6D pose estimation problem as-well as the implementations of proposed pipelines in the real-world. To use the proposed pipelines in the real-world a data set needed to be capture and annotated to train and test the methods. Some controlling robot routines and interfaces were developed in order to be able to control a UR3 robot in the pipelines. The MaskedFusion method, proposed by us, achieves pose estimation accuracy below 6mm in the LineMOD dataset and an AUC score of 93.3% in the challenging YCB-Video dataset. Despite longer training time, MaskedFusion demonstrates low inference time, making it suitable for real-time applications. A study was performed about the effectiveness of employing different color spaces and improved segmentation algorithms to enhance the accuracy of 6D pose estimation methods. Moreover, the proposed MPF6D outperforms other approaches, achieving remarkable accuracy of 99.7% in the LineMOD dataset and 98.06% in the YCB-Video dataset, showcasing its potential for high-precision 6D pose estimation. Additionally, the thesis presents object grasping methods with exceptional accuracy. The first approach, comprising data capture, object detection, 6D pose estimation, grasping detection, robot planning, and motion execution, achieves a 90% success rate in non-controlled environment tests. Leveraging a diverse dataset with varying light conditions proves critical for accurate performance in real-world scenarios. Furthermore, an alternative method demonstrates accurate object grasping without relying on 6D pose estimation, offering faster execution and requiring less computational power. With a remarkable 96% accuracy and an average execution time of 5.59 seconds on a laptop without an NVIDIA GPU, this method demonstrates efficiency and practicality performing unconstrained pick-and-place tasks using a UR3 robot.
- Improving Neural Architecture Search With Bayesian Optimization and Generalization MechanismsPublication . Lopes, Vasco Ferrinho; Alexandre, Luís Filipe Barbosa de AlmeidaAdvances in Artificial Intelligence (AI) and Machine Learning (ML) obtained impressive breakthroughs and remarkable results in various problems. These advances can be largely attributed to deep learning algorithms, especially Convolutional Neural Networks (CNNs). The ever-growing success of CNNs is mainly due to the ingenuity and engineering efforts of human experts who have designed and optimized powerful neural network architectures, which obtained unprecedented results in a vast panoply of tasks. However, applying a ML method to a problem for which it has not been explicitly tailor-made usually leads to sub-optimal results, which in extreme cases can even lead to poor performances, thus hindering the sustainability of a system and the wide-spread application of ML by non-experts. Designing tailor-made CNNs for specific problems is a difficult task, as many design choices depend on each other. Thus, it became logical to automate this process by designing and developing automated Neural Architecture Search (NAS) methods. Architectures found with NAS achieve state-of-the-art performance in various tasks, outperforming human-designed networks. However, NAS methods still face several problems. Most heavily rely on human-defined assumptions constraining the search, such as the architecture’s outer-skeletons, number of layers, parameter heuristics, and search spaces. Common search spaces consist of repeatable modules (cells) instead of fully exploring the architecture’s search space by designing entire architectures (macro-search), which requires deep human expertise and restricts the search to pre-defined settings and narrows the exploration of new and diverse architectures by having forced rules. Also, considerable computation is still inherent to most NAS methods, and only a few can perform macro-search. In this thesis, we focused on proposing novel solutions to mitigate the problems mentioned above. First, we provide a comprehensive review of NAS components, methods, and benchmarks. For the latter, we conduct a study on operation importance to evaluate how the operation pool of search spaces influences the performance of generated architectures. Following, we studied how different neural networks behave for different classification problems and proposed two novel methods to improve upon existing neural networks with NAS by i) searching for a new classification head and ii) searching for a fusion method that allows performing multimodal classification. We then looked into improving the search cost of NAS methods by proposing a zero-proxy estimation strategy that scores architectures at initialization stage through the analysis of the Jacobian matrix and an evolutionary strategy that generates architectures by performing operation mutation and by leveraging the zero-cost proxy estimation to efficiently guide the search process. To further improve the capabilities of NAS methods, we extend the analysis of architectures at initialization stage by proposing a second zero-cost proxy method, which looks at the Neural Tangent Kernel of a generated architecture to infer its final performance if trained. With this, we also propose a novel search space that leverages large pre-trained feature extractors (CNNs) and forces the search only to a small middleware architecture that learns a downstream task. These two methods showed that large models can be efficiently leveraged to learn new tasks without requiring any fine-tuning or extensive computational resources. To further improve the search and memory costs of NAS methods, we proposed MANAS. This method frames NAS as a multi-agent optimization problem and uses independent agents that search for operations in a distributed manner. With MANAS, we showed that both the search cost and the memory resources can be heavily reduced while improving the final performance. Finally, to push NAS to less constrained search spaces and settings, we proposed LCMNAS, a NAS method that performs macrosearch without relying on pre-defined heuristics or bounded search spaces. LCMNAS introduces three components for the NAS pipeline: i) a method that leverages information about well-known architectures to autonomously generate complex search spaces based on weighted directed graphs with hidden properties, ii) an evolutionary search strategy that generates complete architectures from scratch, and iii) a mixed-performance estimation approach that combines information about architectures at initialization stage and lower fidelity estimates to infer their trainability and capacity to model complex functions. Results obtained by the proposed methods show that it is possible to improve NAS methods regarding search and memory costs, as well as computation requirements, while still obtaining state-of-the-art results. All proposed methods were evaluated in multiple search spaces and several data sets, showing improved performances while requiring only a fraction of previous NAS methods’ time and computation needs.
- Efeitos agudos da estimulação transcraniana de corrente contínua no desempenho cognitivo, psicológico e na regulação autonómica de atletas de E-SportsPublication . Machado, Sergio Eduardo de Carvalho; Monteiro, Diogo Manuel Teixeira; Travassos, Bruno Filipe RamaA prática de desporto envolve capacidades fisicas, técnicas e psicológicas, que influenciam diretamente o desempenho dos indivíduos. Uma competição é considerada uma situação stressante, pois provoca alterações fisiológicas, e emocionais nos atletas, como na variabilidade da frequência cardíaca (VFC), perceção de stress e ansiedade competitiva, e consequentemente influenciam o desempenho. Nos desportos tradicionais, há uma maior dependência do desenvolvimento e desempenho tanto das habilidades motoras quanto cognitivas, diferentemente dos desportos eletrónicos (eSports). Os jogadores de eSports dependem muito mais das habilidades cognitivas para o sucesso, daí o motivo de serem conhecidos como “atletas cognitivos”. No entanto, pouco se sabe sobre estratégias eficazes para desenvolver e otimizar o desempenho cognitivo, estado psicológico (perceção de stress e ansiedade cognitiva) e comportamento da VFC em jogadores profissionais de eSports. De forma a alcançar este objetivo, foi realizada uma revisão da literatura, dois estudos originais e um estudo de caso. Os resultados dos quatro estudos científicos englobados nesta tese de doutoramento evidenciaram que: i) a revisão mostrou que a literatura ainda é escassa relativamente com ao uso da estimulação transcraniana de corrente contínua (ETCC) nos eSports, assim como trouxe a hipótese de se utilizar a ETCC como um potencial recurso para aprimorar o funcionamento cognitivo de jogadores profissionais de eSports; ii) os estudos originais demonstraram que o resultado de um jogo de playoff influencia na perceção de stress, na ansiedade competitiva e no comportamento da VFC de jogadores profissioais de eSports; iii) e o estudo de caso sugeriu que a ETCC anódica aplicada ao córtex pré-frontal dorsolateral esquerdo à 2mA reduziu o stress e ansiedade e revelou ainda um aumentou da autoconfiança e o desvio padrão da média do intervalo NN qualificado (SDNN) no momento pós-ETCC comparado aos momentos LB, pré-ETCC e pós-jogo em um jogador profissional de eSports. Em suma, embora os resultados sejam exploratórios, esta tese contribuiu com importantes informações que podem ajudar na avaliação, monitorização, recuperação e melhoria no desempenho de jogadores profissionais de eSports nos aspetos cognitivos, psicológicos e regulação autonómica em treinamentos e jogos.
- Industrial Sensors Online Monitoring and Calibration Through Hidden Markov ModelsPublication . Martins, Alexandre Daniel Batista; Cardoso, António João Marques; Farinha, José Manuel TorresThis thesis aims to demonstrate a methodology able to diagnosis, through the Hidden Markov Model (HMM), the health state of production equipment, as well as the calibration state of sensors reading equipment. Through a well-defined methodology, the observations collected by the sensors are optimised to give input into a HMM, that are translated into hidden states, which represent the diagnosis of the equipment under study, being: State 1 - "Good working"; State 2 - "Warning"; State 3 - "Fault/Uncalibrated". After collecting the data, it goes through a cleaning process that will improve its quality and integrity. Then, a feature generation phase is performed. This phase is extremely important because the information can be managed for the desired equipment. It is through this stage that we can distinguish the diagnosis between the production equipment and the reading equipment. Next, a dimensional reduction of the data is performed, through Principal Component Analysis (PCA) and an extraction of new features that, although in smaller amounts, have more information each one. Then, the new data matrix is applied to a Clustering, performed by K-means, with the objective of grouping similar data within the same group. This will cause good working data to be in one cluster and bad working data to be in a different cluster. These clusters will be the optimized observable states that give input to the HMM. Subsequently, the HMM translates the observable states into a sequence of hidden states that represent the diagnosis of the equipment. Besides the methodology available to detect different types of information from the same data set, it has more capabilities, such as: imputing values in time series with few samples through Deep Neural Network (DNN) methods, namely the Multi-Layer Perceptron (MLP) model; performing the equipment health status prognosis through the Deep Neural Network (DNN), the Gated Recurrent Unit (GRU).
- Hyaluronic acid-functionalized graphene-based nanohybrids for targeted breast cancer chemo-photothermal therapyPublication . Lima-Sousa, Rita; Melo, Bruna L.; Mendonça, António; Correia, I.J.; Melo-Diogo, Duarte deNanomaterials’ application in cancer therapy has been driven by their ability to encapsulate chemotherapeutic drugs as well as to reach the tumor site. Nevertheless, nanomedicines’ translation has been limited due to their lack of specificity towards cancer cells. Although the nanomaterials’ surface can be coated with targeting ligands, such has been mostly achieved through non-covalent functionalization strategies that are prone to premature detachment. Notwithstanding, cancer cells often establish resistance mechanisms that impair the effect of the loaded drugs. This bottleneck may be addressed by using near-infrared (NIR)-light responsive nanomaterials. The NIR-light triggered hyperthermic effect generated by these nanomaterials can cause irreversible damage to cancer cells or sensitize them to chemotherapeutics’ action. Herein, a novel covalently functionalized targeted NIR-absorbing nanomaterial for cancer chemo-photothermal therapy was developed. For such, dopamine-reduced graphene oxide nanomaterials were covalently bonded with hyaluronic acid, and then loaded with doxorubicin (DOX/HA-DOPA-rGO). The produced nanomaterials showed suitable physicochemical properties, high encapsulation efficiency, and photothermal capacity. The in vitro studies revealed that the nanomaterials are cytocompatible and that display an improved uptake by the CD44-overexpressing breast cancer cells. Importantly, the combination of DOX/HA-DOPA-rGO with NIR light reduced breast cancer cells’ viability to just 23 %, showcasing their potential chemo-photothermal therapy.
- The impact of revolutionary aircraft designs on global aviation emissionsPublication . Abrantes, Ivo; Ferreira, Ana F.; Magalhães, Leandro; Costa, Mário; Silva, AndréThe discussion about the environmental impact caused by aviation has gained greater prominence due to the increased demand for this sector and, consequently, the increase in the number of flights. Environmental concerns have stimulated the development of novel approaches to reduce pollutants and CO2 emissions. This study aims to assess the impact of disruptive concepts on commercial aircraft by reducing CO2 emissions by 50% by 2050. In this regard the fleet system dynamics model is used to assess the effects of technological progress on future air transport systems. It accounts for the manufacturer’s production capabilities and current projections and forecasts on the needs and evolution of global air transport, as well as their expected entry into service. The main factors reported were production capacity, year of entry of the technology/concept, and the transport capacity and range of aircraft. The sensitivity study on the production capacity of new aircraft/concepts showed that with a 15% increase, emissions can be reduced between 1 and 2.6%, depending on the case and scenario. On the other hand, increasing the aircraft production capacity could lead to a problem of overcapacity.
- Propolis Protects GC-1spg Spermatogonial Cells against Tert-Butyl Hydroperoxide-Induced Oxidative DamagePublication . Duarte, Filipa Maia; Feijó, Mariana; Luís, Ângelo; Socorro, Sílvia; Maia, Cláudio J.; Correia, SaraPropolis is a natural resin produced by honeybees with plenty of pharmacologic properties, including antioxidant activity. Oxidative stress disrupts germ cell development and sperm function, with demonstrated harmful effects on male reproduction. Several natural antioxidants have been shown to reduce oxidative damage and increase sperm fertility potential; however, little is known about the effects of propolis. This work evaluated the role of propolis in protecting spermatogonial cells from oxidative damage. Propolis’ phytochemical composition and antioxidant potential were determined, and mouse GC-1spg spermatogonial cells were treated with 0.1–500 µg/mL propolis (12–48 h) in the presence or absence of an oxidant stimulus (tert-butyl hydroperoxide, TBHP, 0.005–3.6 µg/mL, 12 h). Cytotoxicity was assessed by MTT assays and proliferation by Ki-67 immunocytochemistry. Apoptosis, reactive oxygen species (ROS), and antioxidant defenses were evaluated colorimetrically. Propolis presented high phenolic and flavonoid content and moderate antioxidant activity, increasing the viability of GC-1spg cells and counteracting TBHP’s effects on viability and proliferation. Additionally, propolis reduced ROS levels in GC-1spg, regardless of the presence of TBHP. Propolis decreased caspase-3 and increased glutathione peroxidase activity in TBHP-treated GC-1spg cells. The present study shows the protective action of propolis against oxidative damage in spermatogonia, opening the possibility of exploiting its benefits to male fertility.
- Flapping Airfoil Aerodynamics using Recurrent Neural NetworkPublication . Pereira, João A.; Camacho, Emanuel A. R.; Marques, Flávio D.; Silva, AndréThe recent increase in interest in artificial intelligence and neural networks has stirred up various industries. Inevitably, its application will trickle down to the most fundamental studies, for instance, unsteady aerodynamics. The present paper serves the purpose of exploring the ability of a recurrent neural network to predict flapping airfoil aerodynamics, in particular the lift coefficient of a plunging NACA0012 airfoil. Thus, a neural network is designed and trained using motion parameters, such as motion frequency and effective angle of attack, to output the instantaneous lift coefficient over a plunging period. Training data is generated using a panel code (HSPM) for fast generation and early testing. Results show that the neural network can adequately predict the lift coefficient for various conditions, including plunging kinematics that are far from the training domain. Future work will build on this framework and extend it to other aerodynamic coefficients using CFD results and experiments, which should enhance the value of the estimates.
- Reduced graphene oxide–reinforced tricalcium phosphate/gelatin/chitosan light-responsive scaffolds for application in bone regenerationPublication . Cabral, Cátia S. D.; Melo-Diogo, Duarte de; Ferreira, Paula; Moreira, André F.; Correia, I.J.Bone is a mineralized tissue with the intrinsic capacity for constant remodeling. Rapid prototyping techniques, using biomaterials that mimic the bone native matrix, have been used to develop osteoinductive and osteogenic personalized 3D structures, which can be further combined with drug delivery and phototherapy. Herein, a Fab@Home 3D Plotter printer was used to promote the layer-by-layer deposition of a composite mixture of gelatin, chitosan, tricalcium phosphate, and reduced graphene oxide (rGO). The phototherapeutic potential of the new NIR-responsive 3D_rGO scaffolds was assessed by comparing scaffolds with different rGO concentrations (1, 2, and 4 mg/mL). The data obtained show that the rGO incorporation confers to the scaffolds the capacity to interact with NIR light and induce a hyperthermy effect, with a maximum temperature increase of 16.7 °C after under NIR irradiation (10 min). Also, the increase in the rGO content improved the hydrophilicity and mechanical resistance of the scaffolds, particularly in the 3D_rGO4. Furthermore, the rGO could confer an NIR-triggered antibacterial effect to the 3D scaffolds, without compromising the osteoblasts' proliferation and viability. In general, the obtained data support the development of 3D_rGO for being applied as temporary scaffolds supporting the new bone tissue formation and avoiding the establishment of bacterial infections.