21 |
[en] COLLABORATIVE FACE TRACKING: A FRAMEWORK FOR THE LONG-TERM FACE TRACKING / [pt] RASTREAMENTO DE FACES COLABORATIVO: UMA METODOLOGIA PARA O RASTREAMENTO DE FACES AO LONGO PRAZOVICTOR HUGO AYMA QUIRITA 22 March 2021 (has links)
[pt] O rastreamento visual é uma etapa essencial em diversas aplicações
de visão computacional. Em particular, o rastreamento facial é considerado
uma tarefa desafiadora devido às variações na aparência da face, devidas
à etnia, gênero, presença de bigode ou barba e cosméticos, além de variações
na aparência ao longo da sequência de vídeo, como deformações,
variações em iluminação, movimentos abruptos e oclusões. Geralmente, os
rastreadores são robustos a alguns destes fatores, porém não alcançam resultados
satisfatórios ao lidar com múltiplos fatores ao mesmo tempo. Uma
alternativa é combinar as respostas de diferentes rastreadores para alcançar
resultados mais robustos. Este trabalho se insere neste contexto e propõe
um novo método para a fusão de rastreadores escalável, robusto, preciso
e capaz de manipular rastreadores independentemente de seus modelos. O
método prevê ainda a integração de detectores de faces ao modelo de fusão
de forma a aumentar a acurácia do rastreamento. O método proposto foi
implementado para fins de validação, tendo sido testado em diversas configurações
que combinaram até cinco rastreadores distintos e um detector de
faces. Em testes realizados a partir de quatro sequências de vídeo que apresentam
condições diversas de imageamento o método superou em acurácia
os rastreadores utilizados individualmente. / [en] Visual tracking is fundamental in several computer vision applications.
In particular, face tracking is challenging because of the variations in facial
appearance, due to age, ethnicity, gender, facial hair, and cosmetics, as well
as appearance variations in long video sequences caused by facial deformations,
lighting conditions, abrupt movements, and occlusions. Generally,
trackers are robust to some of these factors but do not achieve satisfactory
results when dealing with combined occurrences. An alternative is to combine
the results of different trackers to achieve more robust outcomes. This
work fits into this context and proposes a new method for scalable, robust
and accurate tracker fusion able to combine trackers regardless of their models.
The method further provides the integration of face detectors into the
fusion model to increase the tracking accuracy. The proposed method was
implemented for validation purposes and was tested in different configurations
that combined up to five different trackers and one face detector. In
tests on four video sequences that present different imaging conditions the
method outperformed the trackers used individually.
|
22 |
Modelagem e controle de posição e orientação de uma Plataforma de Stewart / Modeling and position and orientation control of a Stewart platformMontezuma, Marcio Aurelio Furtado 14 April 2003 (has links)
Este trabalho apresenta o controle de posição e orientação de um modelo não linear de Plataforma de Stewart com seis graus de liberdade construído no ambiente de sistemas multicorpos ADAMS® desenvolvido pela Mechanical Dynamics, Inc. O modelo não linear é exportado para o ambiente SIMULINK® desenvolvido pela MathWorks, Inc., onde o controle de posição e orientação é realizado a partir da linearização do modelo e a aplicação de um sistema seguidor com realimentação de estados. Utililiza-se, também o SIMULINK® para implementar a dinâmica de um sistema servoválvula e cilindro hidráulico com um servocontrole de pressão e assim simular o comportamento dinâmico de um simulador de vôo com acionamento hidráulico. A utilização destes pacotes comerciais visa obter uma economia de tempo e esforço na modelagem de sistemas mecânicos complexos e na programação para obtenção da resposta do sistema no tempo, além de facilitar a análise de várias configurações de Plataformas de Stewart / This work shows the position and orientation control of a non-linear model of a Stewart platform with six degree of freedom developed in the multibody systems environment ADAMS® (Mechanical Dynamics, Inc.). The non-linear model is exported to SIMULINK® (MathWorks, Inc.), where the position and orientation control is accomplished using the linear model and applied as a tracking-system with state-feedback. The SIMULINK® is also used to implement the dinamics of the servovalve and hydraulic cylinder with pressure-control servo and so simulate the dynamic behavior of the flight simulator with hydraulic drive. These commercial packages are used seeking to save time and effort in the modeling of complex mechanical systems and in the programming to get the time response of the system, facilitating the analysis of several Stewart Platforms configurations.
|
23 |
[en] 3D COLORIZING FOR 2D ANIMATION / [pt] COLORIZAÇÃO 3D PARA ANIMAÇÃO 2DHEDLENA MARIA DE ALMEIDA BEZERRA 31 January 2006 (has links)
[pt] Esta dissertação discute a aplicação de efeitos de
colorização 3D a animações 2D produzidas pela técnica
quadro-a-quadro. Utilizando algoritmos de processamento
de
imagens, desenhos 2D são preparados para receber
técnicas
de sombreamento evitando a transformação da cena para
uma
geometria 3D. Esta preparação se dá através da obtenção
de
mapas de normais que aproximam a geometria do desenho. O
sombreamento é obtido através de um conjunto de técnicas
de renderização foto-realistas e não-foto-realistas, que
podem ser adaptadas para utilização de normais
aproximadas. Visando amenizar o trabalho exaustivo de
colorir cada desenho, um método baseado no
relacionamento
entre imagens é apresentado para colorir automaticamente
cada quadro numa seqüência de desenhos. Este processo de
colorização considera a necessidade de possíveis
intervenções humanas para garantir a qualidade final de
cada imagem da animação. Um estudo sobre aproximação de
normais, técnicas de sombreamento, segmentação de
imagens
e rastreamento de objetos é amplamente discutido nesta
dissertação. / [en] This dissertation discusses the 3D colorization effects
usage over a 2d animation, which has been produced through
frame-by-frame techniques. Normal vector maps approximates
the drawing geometry and provide the ability to perform
shading effects by applying digital image processing
algorithms, avoiding 3D geometry scene transformation. A
set of photorealistic and non-photorealistic renderization
techniques, which can be adapted to normal approximation
usage, is proposed in the colorization process. Also, a
method based on interframe dependence is presented, aiming
to reduce the thoroughgoing effort of colorizing each
individual frame within an animation. This colorization
process considers possible human interventions to ensure
image´s result quality. Finally, this dissertation
provides a comprehensive study regarding several topics,
such as normal approximations, shading techniques, image
segmentation and object tracking.
|
24 |
Ultrafast indoor optical wireless communicationsDiaz, Ariel Gomez January 2016 (has links)
Traffic from wireless and mobile devices is predicted to increase 10-fold between 2014 and 2019, surpassing wired data traffic by 2016. Given the expected radio frequency (RF) capacity crunch, this growing wireless demand will have to be met using a variety of new technologies exploiting other parts of the electromagnetic spectrum. Promising research areas include the Millimetre Band as well as Optical Wireless Communications (OWC). Millimetre Band demonstrations have accomplished ultrafast multi-Gigabit links, making use of state-of-the-art fibre transmission systems. However, complex opto-electronic (OE) interfaces are required to convert the optical carrier into Millimetre wireless signals. To avoid these interfaces, an all-optical transparent network is proposed here, spanning over both the fibre and OWC domains, in order to deliver ultrahigh data rates to mobile end-users in indoor environments. This is supported by the recent deployment of fibre-to-the-home (FTTH) networks creating the potential for Terabit aggregate connections at the user's doorstep. Therefore, infrared fibre-wireless-fibre (FWF) links are studied to support data rates over 100 Gb/s in nomadic applications. The link coverage is achieved via narrow beam beamsteering over a wide field-of-view (FOV) using suitable localization and tracking techniques. The proposed model is inherently bidirectional and transparent, i.e. independent of the data rate and modulation format. In this thesis, the potential for ultrafast wide coverage OWCs using SMF-based transceivers and coherent transmission is demonstrated. A record data rate of 418 Gb/s and 209 Gb/s with a wide FOV of θ<sub>FOV</sub>=±30° and θ<sub>FOV</sub>=±20°, respectively, is shown at a free space range of 3 m. To the best of our knowledge, this is the fastest demonstration of an indoor wireless link that offers practical room-scale coverage. The automated alignment of this FWF link is also demonstrated with the design and implementation a mm-accurate localization and tracking system. Finally, architectures for point-to-multipoint communications are explored in order to adapt the system to multiple users.
|
25 |
Design of information tree for support related queries: Axis Communications AB : An exploratory research study in debug suggestions with machine learning at Axis Communications, Lund / Utformning av informationsträd för supportrelaterade frågor: Axis Communications AB : En utforskande forskningsstudie i felsökningsförslag med maskininlärning vid Axis Communications, LundRangavajjula, Santosh Bharadwaj January 2017 (has links)
Context: In today's world, we have access to so much data than at any time in the past with more and more data coming from smartphones, sensors networks, and business processes. But, most of this data is meaningless, if it's not properly formatted and utilized. Traditionally, in service support teams, issues raised by customers are processed locally, made reports and sent over in the support line for resolution. The resolution of the issue then depends on the expertise of the technicians or developers and their experience in handling similar issues which limits the size, speed, and scale of the problems that can be resolved. One solution to this problem is to make relevant information tailored to the issue under investigation to be easily available. Objectives: The focus of the thesis is to improve turn around time of customer queries using recommendations and evaluate by defining metrics in comparison to existing workflow. As Artificial Intelligence applications can have a broad spectrum, we confine the scope with a relevance in software service and Issue Tracking Systems. Software support is a complicated process as it involves various stakeholders with conflicting interests. During the course of this literary work, we are primarily interested in evaluating different AI solutions specifically in the customer support space customize and compare them. Methods: The following thesis work has been carried out by making controlled experiments using different datasets and Machine learning models. Results: We classified Axis data and Bugzilla (eclipse) using Decision Trees, K Nearest Neighbors, Neural Networks, Naive Bayes and evaluated them using precision, recall rate, and F-score. K Nearest Neighbors was having precision 0.11, recall rate 0.11, Decision Trees had precision 0.11, recall rate 0.11, Neural Networks had precision 0.13, recall rate 0.11 and Naive Bayes had precision 0.05, recall rate 0.11. The result shows too many false positives and true negatives for being able to recommend. Conclusions: In this Thesis work, we have gone through 33 research articles and synthesized them. Existing systems in place and the current state of the art is described. A debug suggestion tool was developed in python with SKlearn. Experiments with different Machine Learning models are run on the tool and highest 0.13 (precision), 0.10 (f-score), 0.11 (recall) are observed with MLP Classification Neural Network.
|
26 |
Supporting Real-Time, Automated Evaluation of Difficulty in Manual AssemblySanthi, B January 2017 (has links) (PDF)
Product designers address the costs of assembly during the design process. Design process can be more efficient if assembly issues could be addressed early in its design process. Doing this requires the ability to assess assemblability among others. Assemblability refers to the ease of assembling the final product from its parts. Assemblability evaluation is applied by product designers for quantitatively estimating the degree of difficulty of the assembly. This helps in identifying areas of improvement, so as to reduce process time and production costs. This work focuses on assessing assemblability in a manual assembly and its importance in the earlier phases of design.
Literature contains various methods for assessing assemblability (e.g. Boothroyd-Dewhurst method, the DFA house, Sturges DFA calculator and Sony DFA method). These methods are typically rule-based and their use requires insight and knowledge on the part of the designer. Further, the designer has to interpret and apply them differently in each specific and unique case. Literature also contains methods for ergonomic assessment of manual work and its link with assemblability. Both observation based ergonomic assessment such as RULA, REBA, VIDAR, LUBA and OWAS and instrumentation based ergonomic assessment using electro-goniometer and accelerometer are the techniques reviewed in this thesis for their suitability in assembly assessment. The most recent trend in the area is automation of the evaluation process. This thesis proposes an approach to automated assessment of assemblability using electromagnetic trackers.
In spite of advances in industrial automation, manual assembly tasks continue to be an important feature of many industrial operations. The method proposed in the thesis is for the assessment of assemblability of manual assembly that combines both time and postural analysis. The tool used for the static analysis is called Rapid Upper Limb Assessment (RULA); for dynamic analysis a new method of time analysis is proposed that is based on the ratio of time spent in fine and gross motions carried out in an assembly process. The difficulty of assembly of a series of manual assembly tasks are carried out in a laboratory setting. Then by correlating this assessment with the feedback on the difficulty of these assembly task obtained from the subjects who carried out these tasks.
The remaining work carried out as a part of this thesis is focussed on automating the process of carrying out the above assessment in an automated manner. Suitability of electromagnetic trackers as a means for automated capture of data necessary for executing the proposed assessment method is studied. Electromagnetic trackers have been used to capture postural data for various limbs of the assembly operators. Data from the limbs are then analysed to identify, to which limb movement signifies which sources of difficulty (i.e. reach, visibility, etc.) in assembly; for example reach difficulty is indicated by torso movement. Finally, the thesis proposes as a part of the future work in possible improvement of the assessment method. Also, its application using a virtual reality (VR) platform assesses in ascertaining ease or difficulties in assembly and many.
|
27 |
Modelagem e controle de posição e orientação de uma Plataforma de Stewart / Modeling and position and orientation control of a Stewart platformMarcio Aurelio Furtado Montezuma 14 April 2003 (has links)
Este trabalho apresenta o controle de posição e orientação de um modelo não linear de Plataforma de Stewart com seis graus de liberdade construído no ambiente de sistemas multicorpos ADAMS® desenvolvido pela Mechanical Dynamics, Inc. O modelo não linear é exportado para o ambiente SIMULINK® desenvolvido pela MathWorks, Inc., onde o controle de posição e orientação é realizado a partir da linearização do modelo e a aplicação de um sistema seguidor com realimentação de estados. Utililiza-se, também o SIMULINK® para implementar a dinâmica de um sistema servoválvula e cilindro hidráulico com um servocontrole de pressão e assim simular o comportamento dinâmico de um simulador de vôo com acionamento hidráulico. A utilização destes pacotes comerciais visa obter uma economia de tempo e esforço na modelagem de sistemas mecânicos complexos e na programação para obtenção da resposta do sistema no tempo, além de facilitar a análise de várias configurações de Plataformas de Stewart / This work shows the position and orientation control of a non-linear model of a Stewart platform with six degree of freedom developed in the multibody systems environment ADAMS® (Mechanical Dynamics, Inc.). The non-linear model is exported to SIMULINK® (MathWorks, Inc.), where the position and orientation control is accomplished using the linear model and applied as a tracking-system with state-feedback. The SIMULINK® is also used to implement the dinamics of the servovalve and hydraulic cylinder with pressure-control servo and so simulate the dynamic behavior of the flight simulator with hydraulic drive. These commercial packages are used seeking to save time and effort in the modeling of complex mechanical systems and in the programming to get the time response of the system, facilitating the analysis of several Stewart Platforms configurations.
|
28 |
Problematika ROI v oblasti získávání a výběru zaměstnanců / The Measures of ROI in recruitmentČevorová, Nina January 2010 (has links)
The paper brings answers on the the question whether it is possible to measure the field of recruitment. Via quantifying the administrative process of recrutment, the paper reflects the savings of time and costs by using e-recruitment application (applicant tracking system ATS) and it searchs for the relationship between quality of hire and increase of company value and its profit.
|
29 |
Decentralized control of multi-agent systems : a hybrid formalism / Commande décentralisée de systèmes multi agents : un formalisme hybrideBorzone, Tommaso 09 September 2019 (has links)
Au cours des dernières années, les problèmes multi-agents ont été étudiés de manière intensive par la communauté de la théorie du contrôle. L'un des sujets les plus populaires est le problème de consensus où un groupe d'agents parvient à un accord sur la valeur d'un certain paramètre ou d’une variable. Dans ce travail, nous nous concentrons sur le consensus des réseaux d'agents avec une dynamique non linéaire de poursuite de référence. Nous utilisons des interactions sporadiques modélisées par la détection relative, pour traiter le consensus décentralisé des références. La référence est donc utilisée pour alimenter la dynamique de poursuite de chaque agent. L'analyse de stabilité du système globale a nécessitée l'utilisation d'outils théoriques propre de la théorie des systèmes hybrides, en raison de la double nature de l'approche en deux étapes. L'analyse est effectuée en tenant compte de différents scénarios de topologie et interactions. Pour chaque cas, une condition suffisante de stabilité est fournie, en termes de temps minimum autorisé entre deux mises à jour de référence consécutives. Le cadre proposé est appliqué aux missions de rendez-vous et de réalisation de formation pour les robots mobiles non-holonomes. Le même problème est abordé dans le contexte d'une application réelle sur le terrain, à savoir un système de gestion de flotte pour un groupe de véhicules robotisés déployés dans un environnement industriel à des fins de surveillance et de collecte de données. Le développement d'une telle application a été motivé par le fait que cette thèse s'inscrit dans le cadre du projet FFLOR, développé par le département de recherche technologique du CEA tech. / Over the last years, multi-agents problems have been extensively studied from the control theory community. One of the most popular multi-agents control topics is the consensus problem where a group of agents reaches an agreement over the value of a certain parameter or variable. In this work we focus our attention on the consensus problem of networks of non-linear reference tracking agents. In first place, we use sporadic interactions modeled by relative sensing to deal with the decentralized consensus of the references. The reference is therefore feeded the tracking dynamics of each agent. Differently from existent works, the stability analysis of the overall system required the usage of hybrid systems theory tools, due to dual nature of the two stages approach. The analysis is carried out considering different scenarios of network topology and interactions. For each case a stability sufficient condition in terms of the minimum allowed time between two consecutive reference updates is provided. The proposed framework is applied to the rendez-vous and formation realisation tasks for non-holonomic mobile robots, which appear among the richest research topics in recent years. The same problem is addressed in the context of a real field application, namely a fleet management system for a group of robotic vehicles deployable in an industrial environment for monitoring and data collection purpose. The development of such application was motivated by the fact that this thesis is part of the Future of Factory Lorraine (FFLOR) project, developed by the technological research department of the Commissariat à l'énergie atomique et aux énergies alternatives (CEA tech).
|
30 |
Electronic Performance And Tracking Systems (EPTS) : Perceptions, Benefits and Challenges of Professional Football Athletes and Training StaffBitilis, Pavlos January 2021 (has links)
Personal Informatics (PI) are information systems that allow people to process activities with the usage of information technology, aiming to produce informational products (data) either for themselves or for others. Technologies that enable PI are becoming increasingly popular, assisting people in collecting personally relevant information about their body and their behaviour. In sports industry nowadays, a great variety of PI wearable tools offer support to athletes and training staff to improve their performance. An example of such tool is the Electronic Performance and Tracking Systems (EPTS), which are a combination of hardware and software that facilitates the collection, storage, analysis and management of professional athletes’ fitness and health data. Although significant and broadly used, EPTS have not yet received much attention from researchers and, thus, understudied. Therefore, the master’s thesis explores the perceptions of professional football athletes and training staff regarding the use of EPTS in their everyday training and work. Furthermore, the master’s thesis research explores the benefits and challenges that professional football athletes and training staff experience when using EPTS in their everyday training and work. The master’s thesis study adopts the interpretive paradigm and qualitative ethnographic approach. The research data was collectedthrough direct observations in the field and semi-structured interviews from Greek professional football athletes and Greek training staff that use wearable EPTS in their everyday training and work and was analysed thematically. A theoretical framework, which is built upon relevant literature from the informatics field and along with the theory of sensemaking, is used to understand, interpret and discuss the research findings. The research outcome of the master’s thesis shows that communication is at the core of EPTS enabling football players and training staff to improve individual and team performance. Organizing of every day starts and ends with EPTS analysis and evaluation and better organized and daily evaluated football methodology appears as key benefit for the club. Coaches and trainers are now more data driven and accurate and analysts and trainers that conduct analysisof the data provided by EPTS are new members of the training staff. Evidence provided by EPTS build trust between staff and players and in the training staff. Visualization tools for presenting insights need to be further improved with the addition of in-field monitors and 3D presentations. Furthermore, it is important for training staff members to have ethical and consistent strategy on how data derived from EPTS are used on how data are communicated. The research complements previous research on personal informatics and adjusts them to elite team sport context and adds to the theory of sensemaking regarding how users make sense of PI tools that are related with their everyday routines at work. In addition, it offers football training staff members a model for efficient use of EPTS technology into the everyday football practices and a model of sustainable use aiming the overall improvement of team performance.
|
Page generated in 0.0842 seconds