A contribution to semantic description of images and videos: an application of soft biometrics

Humans have a high ability to extract visual data information acquired by sight. Trought a learning process, which starts at birth and continues throughout life, image interpretation becomes almost instinctively. At a glance, one can easily describe a scene with reasonable precision, naming its main components. Usually, this is done by extracting low-level features such as edges, shapes and textures, and associanting them to high level meanings. In this way, a semantic description of the scene is done. An example of this, is the human capacity to recognize and describe other people physical and behavioral characteristics, or biometrics. Soft-biometrics also represents inherent characteristics of human body and behaviour, but do not allow unique person identification. Computer vision area aims to develop methods capable of performing visual interpretation with performance similar to humans. This thesis aims to propose computer vison methods which allows high level information extraction from images in the form of soft biometrics. This problem is approached in two ways, unsupervised and supervised learning methods. The first seeks to group images via an automatic feature extraction learning , using both convolution techniques, evolutionary computing and clustering. In this approach employed images contains faces and people. Second approach employs convolutional neural networks, which have the ability to operate on raw images, learning both feature extraction and classification processes. Here, images are classified according to gender and clothes, divided into upper and lower parts of human body. First approach, when tested with different image datasets obtained an accuracy of approximately 80% for faces and non-faces and 70% for people and non-person. The second tested using images and videos, obtained an accuracy of about 70% for gender, 80% to the upper clothes and 90% to lower clothes. The results of these case studies, show that proposed methods are promising, allowing the realization of automatic high level information image annotation. This opens possibilities for development of applications in diverse areas such as content-based image and video search and automatica video survaillance, reducing human effort in the task of manual annotation and monitoring.

A single theoretical framework for circular features processing in humans: orientation and direction of motion compared

Common computational principles underlie processing of various visual features in the cortex. They are considered to create similar patterns of contextual modulations in behavioral studies for different features as orientation and direction of motion. Here, I studied the possibility that a single theoretical framework, implemented in different visual areas, of circular feature coding and processing could explain these similarities in observations. Stimuli were created that allowed direct comparison of the contextual effects on orientation and motion direction with two different psychophysical probes: changes in weak and strong signal perception. One unique simplified theoretical model of circular feature coding including only inhibitory interactions, and decoding through standard vector average, successfully predicted the similarities in the two domains, while different feature population characteristics explained well the differences in modulation on both experimental probes. These results demonstrate how a single computational principle underlies processing of various features across the cortices.

Obtenção e caracterização de queratinase de bacillus sp. P45 a partir de coprodutos e aplicação na produção de queijos cremosos enriquecidos com chia e quinoa

As proteases constituem 60-65% do mercado global das enzimas industriais e são utilizadas na indústria de alimentos no processo de amaciamento de carne, na síntese de peptídeos, preparo de fórmulas infantis, panificação, cervejarias, produtos farmacêuticos, diagnósticos médicos, como aditivos na indústria de detergentes e na indústria têxtil no processo de depilação e transformação do couro. Proteases específicas produzidas por micro-organismos queratinolíticos são chamadas de queratinases e distinguem-se de outras proteases pela maior capacidade de degradação de substratos compactos e insolúveis como a queratina. Atualmente, processos que apontem o uso total das matérias-primas e que não resultem em impactos negativos ao meio ambiente tem ganhado destaque. Dentro desta temática, destacam-se a reutilização da farinha de penas residual durante o cultivo do Bacillus sp. P45 para produção de proteases e a biomassa residual de levedura, ambas com elevados teores de proteínas, podendo ser utilizadas no cultivo do Bacillus sp. P45 para obtenção de proteases. O objetivo deste trabalho foi obter a enzima queratinase purificada em grandes quantidades, sua caracterização, bem como a sua aplicação em processos de coagulação enzimática do leite para o desenvolvimento de um queijo cremoso enriquecido com farinha de chia e quinoa. Além disso, aplicar diferentes coprodutos para produção de enzimas proteolíticas e queratinolíticas. A presente tese foi dividida em quatro artigos: no primeiro foi realizado a obtenção da queratinase purificada em maiores quantidades e a determinação dos parâmetros de estabilidade térmica e a influência de componentes químicos na atividade enzimática. A obtenção da enzima em maiores quantidades alcançou fatores de purificação de 2,6, 6,7 e 4,0 vezes, paras 1º SAB, 2º SAB e diafiltração, respectivamente. A recuperação enzimática alcançou valores de 75,3% para o 1º SAB, 75,1% no 2º sistema e 84,3% na diafiltração. A temperatura de 55ºC e o pH 7,5 foram determinados como ótimos para atividade da enzima queratinase. O valor da energia de desativação (Ed) médio foi de 118,0 kJ/mol e os valores de z e D variaram de 13,6 a 18,8ºC, e 6,9 a 237,3 min, respectivamente. Além disso a adição de sais (CaCl2, CaO, C8H5KO4 e MgSO4) elevou a atividade da enzima na presença destes compostos. O segundo artigo apresenta a aplicação da queratinase como coagulante de leite bovino e sua aplicação na obtenção de queijo cremoso enriquecido com chia e quinoa. A enzima mostrou atividade de coagulação semelhante ao coagulante comercial, na concentração de 30mg/mL. A enzima purificada foi empregada de forma eficiente na fabricação do queijo cremoso, que apresentou valores de pH de 5,3 e acidez de 0,06 a 0,1 mol/L, com elevação durante os 25 dias de armazenamento. O terceiro artigo apresenta o perfil do queijo cremoso enriquecido com farinha de chia e quinoa, o qual apresentou alto índice de retenção de água (>99,0%) e baixos valores de sinérese (<0,72%). Elevados teores de fibras foi verificado (3,0 a 5,0%), sugerindo seu consumo como fonte de fibras. As análises microbiológicas foram de acordo com a legislação vigente. Na análise sensorial foi verificado altos valores de suavidade ao paladar e verificado maiores valores de consistência e untabilidade nas amostras com maiores concentrações de nata e quinoa. O quarto artigo traz a extração de β-galactosidase por ultrassom e o uso da biomassa residual da levedura, bem como o uso de farinha de penas residuais como substrato para obtenção de proteases. O ultrassom foi eficiente para ruptura celular e extração de β-galactosidase, apresentando alta atividade (35,0 U/mL) e rendimento (876,0 U/g de biomassa). A maior atividade proteolítica (1300 U/mL em 32 h) e queratinolítica (89,2 U/mL) verificadas ocorreram utilizando-se a biomassa e a farinha de penas residuais, respectivamente. Maior produtividade proteolítica (40,8 U/mL/h) foi verificado no meio utilizando biomassa residual como substrato. Já a maior produtividade queratinolítica (2,8 U/mL/h) foi alcançada utilizando farinha de penas reutilizada.

DORIS: Processing and management of Sound Velocity Profiles for echosounding applications

If a bathymetric echosounder is the essential device to carry on hydrographic surveys, other external sensors are absolutely also necessary (positioning system, motion unit or sound velocity profiler). And because sound doesn‛t go straight away into the whole bathymetric swath its measurement and processing are very sensitive for all the water column. DORIS is the very answer for an operational sound velocity profile processing.

Off-line and On-line Affective Recognition of a Computer User through A Biosignal Processing Approach

Physiological signals, which are controlled by the autonomic nervous system (ANS), could be used to detect the affective state of computer users and therefore find applications in medicine and engineering. The Pupil Diameter (PD) seems to provide a strong indication of the affective state, as found by previous research, but it has not been investigated fully yet. In this study, new approaches based on monitoring and processing the PD signal for off-line and on-line affective assessment (“relaxation” vs. “stress”) are proposed. Wavelet denoising and Kalman filtering methods are first used to remove abrupt changes in the raw Pupil Diameter (PD) signal. Then three features (PDmean, PDmax and PDWalsh) are extracted from the preprocessed PD signal for the affective state classification. In order to select more relevant and reliable physiological data for further analysis, two types of data selection methods are applied, which are based on the paired t-test and subject self-evaluation, respectively. In addition, five different kinds of the classifiers are implemented on the selected data, which achieve average accuracies up to 86.43% and 87.20%, respectively. Finally, the receiver operating characteristic (ROC) curve is utilized to investigate the discriminating potential of each individual feature by evaluation of the area under the ROC curve, which reaches values above 0.90. For the on-line affective assessment, a hard threshold is implemented first in order to remove the eye blinks from the PD signal and then a moving average window is utilized to obtain the representative value PDr for every one-second time interval of PD. There are three main steps for the on-line affective assessment algorithm, which are preparation, feature-based decision voting and affective determination. The final results show that the accuracies are 72.30% and 73.55% for the data subsets, which were respectively chosen using two types of data selection methods (paired t-test and subject self-evaluation). In order to further analyze the efficiency of affective recognition through the PD signal, the Galvanic Skin Response (GSR) was also monitored and processed. The highest affective assessment classification rate obtained from GSR processing is only 63.57% (based on the off-line processing algorithm). The overall results confirm that the PD signal should be considered as one of the most powerful physiological signals to involve in future automated real-time affective recognition systems, especially for detecting the “relaxation” vs. “stress” states.

Effect of room temperature transport vials on DNA quality and phylogenetic composition of faecal microbiota of elderly adults and infants

Background: Alterations in intestinal microbiota have been correlated with a growing number of diseases. Investigating the faecal microbiota is widely used as a non-invasive and ethically simple proxy for intestinal biopsies. There is an urgent need for collection and transport media that would allow faecal sampling at distance from the processing laboratory, obviating the need for same-day DNA extraction recommended by previous studies of freezing and processing methods for stool. We compared the faecal bacterial DNA quality and apparent phylogenetic composition derived using a commercial kit for stool storage and transport (DNA Genotek OMNIgene GUT) with that of freshly extracted samples, 22 from infants and 20 from older adults. Results: Use of the storage vials increased the quality of extracted bacterial DNA by reduction of DNA shearing. When infant and elderly datasets were examined separately, no differences in microbiota composition were observed due to storage. When the two datasets were combined, there was a difference according to a Wilcoxon test in the relative proportions of Faecalibacterium, Sporobacter, Clostridium XVIII, and Clostridium XlVa after 1 week's storage compared to immediately extracted samples. After 2 weeks' storage, Bacteroides abundance was also significantly different, showing an apparent increase from week 1 to week 2. The microbiota composition of infant samples was more affected than that of elderly samples by storage, with significantly higher Spearman distances between paired freshly extracted and stored samples (p

Emerging plant-based proteins: comparison from extraction to functional properties and food application

The increasing demand for alternatives to meat food products, which is linked to ethical and environmental reasons, highlights the necessity of using different protein sources. Plant proteins provide a valid option, thanks to the relative low costs, high availability and wide supply sources. The current process used to produce plant concentrates and isolates is the alkaline extraction followed by isoelectric precipitation. However, despite the high purity of the proteins, it presents some drawbacks. Innovative protein extraction processes are emerging, with the aim of reducing the environmental impact and the costs, as well as improving the functional properties. In this study, the traditional wet protein extraction and another simplified wet process were used to obtain protein-rich extracts out of different plants. The sources considered in the project were de-oiled sunflower and canola, chickpea, lentils, and the camelina meal, an emerging oleaginous seed interesting for its high content of omega 3. The extracts obtained from the two processes were then analysed for their capacities to hold water and fat, to form gel and a stable foam. Results highlighted strong differences concerning the protein content, yield and functionalities. The extracts obtained with the alkaline process confirmed the literature data about the four plant sources (sunflower, canola, chickpea and lentils) and allow to obtain a camelina concentrate with a protein content of 63 % and a protein recovery of 41 %. The second easiest process was not effective to obtain a protein enrichment in oleaginous sources, whereas an enrichment of 10 and 15 % was obtained in chickpea and lentils, respectively. The functional properties were also completely different: the easiest process produced protein ingredients completely water-soluble at pH 7, with a discrete foaming capacity compared to the extracts obtained with alkaline process. These characteristics could make these extracts suitable for the plant milk-analogue products.

Deep and near space tracking stations in support of lunar and planetary exploration missions

The aim of this dissertation is to describe the methodologies required to design, operate, and validate the performance of ground stations dedicated to near and deep space tracking, as well as the models developed to process the signals acquired, from raw data to the output parameters of the orbit determination of spacecraft. This work is framed in the context of lunar and planetary exploration missions by addressing the challenges in receiving and processing radiometric data for radio science investigations and navigation purposes. These challenges include the designing of an appropriate back-end to read, convert and store the antenna voltages, the definition of appropriate methodologies for pre-processing, calibration, and estimation of radiometric data for the extraction of information on the spacecraft state, and the definition and integration of accurate models of the spacecraft dynamics to evaluate the goodness of the recorded signals. Additionally, the experimental design of acquisition strategies to perform direct comparison between ground stations is described and discussed. In particular, the evaluation of the differential performance between stations requires the designing of a dedicated tracking campaign to maximize the overlap of the recorded datasets at the receivers, making it possible to correlate the received signals and isolate the contribution of the ground segment to the noise in the single link. Finally, in support of the methodologies and models presented, results from the validation and design work performed on the Deep Space Network (DSN) affiliated nodes DSS-69 and DSS-17 will also be reported.

IDeS methodology applied to automotive component re-engineering focused on lightweight and high-strength by additive manufacturing

The project aims to gather an understanding of additive manufacturing and other manufacturing 4.0 techniques with an eyesight for industrialization. First the internal material anisotropy of elements created with the most economically feasible FEM technique was established. An understanding of the main drivers for variability for AM was portrayed, with the focus on achieving material internal isotropy. Subsequently, a technique for deposition parameter optimization was presented, further procedure testing was performed following other polymeric materials and composites. A replicability assessment by means of the use of technology 4.0 was proposed, and subsequent industry findings gathered the ultimate need of developing a process that demonstrate how to re-engineer designs in order to show the best results with AM processing. The latest study aims to apply the Industrial Design and Structure Method (IDES) and applying all the knowledge previously stacked into fully reengineer a product with focus of applying tools from 4.0 era, from product feasibility studies, until CAE – FEM analysis and CAM – DfAM. These results would help in making AM and FDM processes a viable option to be combined with composites technologies to achieve a reliable, cost-effective manufacturing method that could also be used for mass market, industry applications.

Integrating the age factor in designing industrial environments and workstations in the workforce ageing era

As people spend a third of their lives at work and, in most cases, indoors, the work environment assumes crucial importance. The continuous and dynamic interaction between people and the working environment surrounding them produces physiological and psychological effects on operators. Recognizing the substantial impact of comfort and well-being on employee satisfaction and job performance, the literature underscores the need for industries to implement indoor environment control strategies to ensure long-term success and profitability. However, managing physical risks (i.e., ergonomic and microclimate) in industrial environments is often constrained by production and energy requirements. In the food processing industry, for example, the safety of perishable products dictates storage temperatures that do not allow for operator comfort. Conversely, warehouses dedicated to non-perishable products often lack cooling systems to limit energy expenditure, reaching high temperatures in the summer period. Moreover, exceptional events, like the COVID-19 pandemic, introduce new constraints, with recommendations impacting thermal stress and respiratory health. Furthermore, the thesis highlights how workers' variables, particularly the aging process, reduce tolerance to environmental stresses. Consequently, prolonged exposure to environmental stress conditions at work results in cardiovascular disease and musculoskeletal disorders. In response to the global trend of an aging workforce, the thesis bridges a literature gap by proposing methods and models that integrate the age factor into comfort assessment. It aims to present technical and technological solutions to mitigate microclimate risks in industrial environments, ultimately seeking innovative ways to enhance the aging workforce's comfort, performance, experience, and skills. The research outlines a logical-conceptual scheme with three main areas of focus: analyzing factors influencing the work environment, recognizing constraints to worker comfort, and designing solutions. The results significantly contribute to science by laying the foundation for new research in worker health and safety in an ageing working population's extremely current industrial context.

Computer vision in aseptic pharmaceutical manufacturing: a deep learning-based approach to vial inspection and pack formation during a lyophilizer loading cycle

Vision systems are powerful tools playing an increasingly important role in modern industry, to detect errors and maintain product standards. With the enlarged availability of affordable industrial cameras, computer vision algorithms have been increasingly applied in industrial manufacturing processes monitoring. Until a few years ago, industrial computer vision applications relied only on ad-hoc algorithms designed for the specific object and acquisition setup being monitored, with a strong focus on co-designing the acquisition and processing pipeline. Deep learning has overcome these limits providing greater flexibility and faster re-configuration. In this work, the process to be inspected consists in vials’ pack formation entering a freeze-dryer, which is a common scenario in pharmaceutical active ingredient packaging lines. To ensure that the machine produces proper packs, a vision system is installed at the entrance of the freeze-dryer to detect eventual anomalies with execution times compatible with the production specifications. Other constraints come from sterility and safety standards required in pharmaceutical manufacturing. This work presents an overview about the production line, with particular focus on the vision system designed, and about all trials conducted to obtain the final performance. Transfer learning, alleviating the requirement for a large number of training data, combined with data augmentation methods, consisting in the generation of synthetic images, were used to effectively increase the performances while reducing the cost of data acquisition and annotation. The proposed vision algorithm is composed by two main subtasks, designed respectively to vials counting and discrepancy detection. The first one was trained on more than 23k vials (about 300 images) and tested on 5k more (about 75 images), whereas 60 training images and 52 testing images were used for the second one.

Sargassum Filipendula Alginate From Brazil: Seasonal Influence And Characteristics.

The aim of this work is focused on the extraction and characterization of the Brazilian seaweed Sargassum filipendula alginate. Alginates obtained at different seasons were characterized by liquid state nuclear magnetic resonance spectroscopy and scanning electron microscopy. The alginate extraction efficiency was about 20%. Different seasons of the year and different stages in the life cycle of Sargassum sp. in southeastern Brazil influenced the M/G and, consequently, the technological properties of extracted alginates.

Development And Characterization Of Microsatellite Loci For Ocotea Species (lauraceae) Threatened With Extinction.

The Atlantic rainforest species Ocotea catharinensis, Ocotea odorifera, and Ocotea porosa have been extensively harvested in the past for timber and oil extraction and are currently listed as threatened due to overexploitation. To investigate the genetic diversity and population structure of these species, we developed 8 polymorphic microsatellite markers for O. odorifera from an enriched microsatellite library by using 2 dinucleotide repeats. The microsatellite markers were tested for cross-amplification in O. catharinensis and O. porosa. The average number of alleles per locus was 10.2, considering all loci over 2 populations of O. odorifera. Observed and expected heterozygosities for O. odorifera ranged from 0.39 to 0.93 and 0.41 to 0.92 across populations, respectively. Cross-amplification of all loci was successfully observed in O. catharinensis and O. porosa except 1 locus that was found to lack polymorphism in O. porosa. Combined probabilities of identity in the studied Ocotea species were very low ranging from 1.0 x 10-24 to 7.7 x 10-24. The probability of exclusion over all loci estimated for O. odorifera indicated a 99.9% chance of correctly excluding a random nonparent individual. The microsatellite markers described in this study have high information content and will be useful for further investigations on genetic diversity within these species and for subsequent conservation purposes.

Genomic And Chemical Insights Into Biosurfactant Production By The Mangrove-derived Strain Bacillus Safensis Ccma-560.

Many Bacillus species can produce biosurfactant, although most of the studies on lipopeptide production by this genus have been focused on Bacillus subtilis. Surfactants are broadly used in pharmaceutical, food and petroleum industry, and biological surfactant shows some advantages over the chemical surfactants, such as less toxicity, production from renewable, cheaper feedstocks and development of novel recombinant hyperproducer strains. This study is aimed to unveil the biosurfactant metabolic pathway and chemical composition in Bacillus safensis strain CCMA-560. The whole genome of the CCMA-560 strain was previously sequenced, and with the aid of bioinformatics tools, its biosurfactant metabolic pathway was compared to other pathways of closely related species. Fourier transform infrared (FTIR) and high-resolution TOF mass spectrometry (MS) were used to characterize the biosurfactant molecule. B. safensis CCMA-560 metabolic pathway is similar to other Bacillus species; however, some differences in amino acid incorporation were observed, and chemical analyses corroborated the genetic results. The strain CCMA-560 harbours two genes flanked by srfAC and srfAD not present in other Bacillus spp., which can be involved in the production of the analogue gramicidin. FTIR and MS showed that B. safensis CCMA-560 produces a mixture of at least four lipopeptides with seven amino acids incorporated and a fatty acid chain with 14 carbons, which makes this molecule similar to the biosurfactant of Bacillus pumilus, namely, pumilacidin. This is the first report on the biosurfactant production by B. safensis, encompassing the investigation of the metabolic pathway and chemical characterization of the biosurfactant molecule.

Uma nova classificação de alimentos baseada na extensão e propósito do seu processamento

Este artigo descreve uma nova classificação de alimentos baseada na extensão e propósito do processamento industrial usado na sua produção. Três grupos são definidos: alimentos não processados ou minimamente processados (grupo 1), alimentos processados utilizados como ingredientes de preparações culinárias ou pela indústria de alimentos (grupo 2), e produtos alimentícios ultra-processados (grupo 3). O uso da classificação é ilustrado aplicando-a a dados coletados por Pesquisa de Orçamentos Familiares conduzida em 2002/2003 em uma amostra probabilística de 48.470 domicílios brasileiros. A disponibilidade diária foi de 1.792kcal/capita, sendo 42,5por cento de alimentos do grupo 1, 37,5por cento do grupo 2 e 20por cento do grupo 3. A contribuição do grupo 3 aumentou com a renda familiar, correspondendo a um terço do total calórico nos domicílios mais afluentes. Discute-se o impacto sobre a qualidade geral da dieta, padrões de alimentação e condições de saúde que poderia ocorrer com a substituição de alimentos do grupo 1 e ingredientes do grupo 2 por produtos alimentícios do grupo 3