DI - CAOS - Capítulos de Monografías

Permanent URI for this collection

https://hdl.handle.net/10016/9756

Browse

Now showing 1 - 20 of 45

Computer vision and laser scanner road environment perception
(IEEE, 2014-05-12) García, Fernando; Ponz Vila, Aurelio; Martín Gómez, David; Escalera Hueso, Arturo de la; Armingol Moreno, José María
Data fusion procedure is presented to enhance classical Advanced Driver Assistance Systems (ADAS). The novel vehicle safety approach, combines two classical sensors: computer vision and laser scanner. Laser scanner algorithm performs detection of vehicles and pedestrians based on pattern matching algorithms. Computer vision approach is based on Haar-Like features for vehicles and Histogram of Oriented Gradients (HOG) features for pedestrians. The high level fusion procedure uses Kalman Filter and Joint Probabilistic Data Association (JPDA) algorithm to provide high level detection. Results proved that by means of data fusion, the performance of the system is enhanced.
Context Aided Multilevel Pedestrian Detection
(IEEE, 2013) García, Fernando; Escalera Hueso, Arturo de la; Armingol Moreno, José María
Abstract: The proposed work, depicts a novel algorithm able to provide multiple pedestrian detection, based on the use of classical sensors in modern automotive application and context information. The work takes advantage of the use of Joint Probabilities Data Association (JPDA) and context information to enhance the classic performance of the pedestrian detection algorithms. The combination of the different information sources with powerful tracking algorithms helps to overcome the difficulties of this processes, providing a trustable tool that improves performance of the single sensor detection algorithms. Context in a rich information source, able to improve the fusion process in all levels by the use of a priori knowledge of the application. In the present work multilevel fusion solution is provided for road safety application. Context is used in all the fusion levels, helping to improve the perception of the road environment and the relations among detections. By the fusion of all information sources, accurate and trustable detection is provided and complete situation assessment obtained, with estimation of the danger that involves any detection.
Joint Probabilistic Data Association fusion approach for pedestrian detection
(Ieee - The Institute Of Electrical And Electronics Engineers, Inc, 2013) García, Fernando; Escalera Hueso, Arturo de la; Armingol Moreno, José María
Abstract: Fusion is becoming a classic topic in Intelligent Transport System (ITS) society. The lack of trustworthy sensors requires the combination of several devices to provide reliable detections. In this paper a novel approach, that takes advantage of the Joint Probabilistic Data Association technique (JPDA) for data association, is presented. The approach uses one of the most powerful techniques of Multiple Target Tracking theory and adapts it to fulfill the strong requirements of road safety applications. The different test performed proved that a powerful association technique can enhance the capacity of Advance Driver Assistance Systems. Two main sensors are used for pedestrian detection: laser scanner and computer vision. Furthermore, the approach takes advantage of the availability of other information sources i.e. context information and online information (GPS). The detections are fused using JPDA, enhancing the capacities of classical pedestrian detection systems, mainly based in visual information. The test performed also showed that JPDA improved the results offered by other data association techniques, e.g. Global Nearest Neighbors.
Ensemble method based on individual evolving classifiers
(IEEE Computer Society, 2013) Iglesias Martínez, José Antonio; Ledezma Espino, Agapito Ismael; Sanchis de Miguel, María Araceli
Abstract: Humans often seek a second or third opinion about an important matter. Then, a final decision is reached after weighing and combining these opinions. This idea is the base of the ensemble based systems. Ensembles of classifiers are well established as a method for obtaining highly accurate classifiers by combining less accurate ones. On the other hand, evolving classifiers are inspired by the idea of evolve their structure in order to adapt to the changes of the environment. In this paper, we present a proof-of-concept method for constructing an ensemble system based on Evolving Fuzzy Systems. The main contribution of this approach is that the base-classifiers are self-developing (evolving) Fuzzy-rule-based (FRB) classifiers. Thus, we present an ensemble system which is based on evolving classifiers and keeps the properties of the evolving approach classification of streaming data. It is important to clarify that the evolving classifiers are gradually developing but they are not genetic or evolutionary.
Evolving Systems for Computer User Behavior Classification
(IEEE, 2013) Iglesias Martínez, José Antonio; Ledezma Espino, Agapito Ismael; Sanchis de Miguel, María Araceli
A computer can keep track of computer users to improve the security in the system. However, this does not prevent a user from impersonating another user. Only the user behavior recognition can help to detect masqueraders. Under the UNIX operating system, users type several commands which can be analyzed in order to create user profiles. These profiles identify a specific user or a specific computer user behavior. In addition, a computer user behavior changes over time. If the behavior recognition is done automatically, these changes need to be taken into account. For this reason, we propose in this paper a simple evolving method that is able to keep up to date the computer user behavior profiles. This method is based on Evolving Fuzzy Systems. The approach is evaluated using real data streams.
Automatic design of artificial neural networks to forecast time series
(Ibergarceta, 2010) Peralta, Juan; Gutiérrez Sánchez, Germán; Sanchis de Miguel, María Araceli
In this work an approach to design Artificial Neural Networks (ANN) to forecast Time Series is tackled. The approach is an automatic method that is carried out by an Evolutionary Algorithm (as a search algorithm) to design ANN. A key issue for these kinds of approaches is what information is included into the chromosome that represents an ANN There are two principal ideas about this question: first, the chromosome contains information about parameters of the topology, architecture, learning parameters, etc. of the ANN. The results using a parameter Encoding Scheme to design ANN for a Time Series Competition are shown
Testing feature selection in traffic signs
(Universidad de Las Palmas de Gran Canaria, 2007) Sesmero Lorente, María Paz; Alonso Weber, Juan Manuel; Gutiérrez Sánchez, Germán; Sanchis de Miguel, María Araceli
Road signs carry essential information for successful driving. Therefore, if we are interested in developing a Driver Support Systems, both, detection and classification of road signs are essential tasks for an autonomous system. However, both tasks are some of the less studied subjects in the field of Intelligent Transport systems. In this research we lay the foundations of a software implementation for a classifier system that will be implemented in hardware and will be able to be used for real-time traffic sign categorization. The selected classification method is a Multilayer Perceptron trained with Back-Propagation algorithm. The reason of this selection is, on one hand, that for certain types of problems, such as object recognition in natural environments, neural network learning methods provide a robust approach. On the other hand, and under certain, limitations related mainly to the number of units, a hardware implementation on FPGA of ANN is possible. Therefore, ANNs are a good method for real-time processing in real-word problems
Cooperación en sistemas distribuidos de robots reactivos minimizando la cantidad de información comunicada
(Universidad de Vigo, 2000) Fernández, Fernando; Gutiérrez, Germán; Molina, José M.
La coordinación emergente pretende obtener comportamientos colaborativos entre diversos agentes sin que eso implique que cada individuo deba tener un conocimiento global del dominio, y sin que ese conocimiento deba estar centralizado. Al no requerir conocimiento global, se minimiza la comunicación entre los agentes de forma que cada uno de ellos puede comportarse de forma reactiva y totalmente autónoma. En este trabajo se presenta una primera aproximación a este modelo de coordinación aplicado al dominio de la RoboCup.
MMRF for proteome annotation applied to human protein disease prediction
(Springer, 2011) García Jiménez, Beatriz; Ledezma Espino, Agapito Ismael; Sanchis de Miguel, María Araceli
Biological processes where every gene and protein participates is an essential knowledge for designing disease treatments. Nowadays, these annotations are still unknown for many genes and proteins. Since making annotations from in-vivo experiments is costly, computational predictors are needed for different kinds of annotation such as metabolic pathway, interaction network, protein family, tissue, disease and so on. Biological data has an intrinsic relational structure, including genes and proteins, which can be grouped by many criteria. This hinders the possibility of finding good hypotheses when attribute-value representation is used. Hence, we propose the generic Modular Multi-Relational Framework (MMRF) to predict different kinds of gene and protein annotation using Relational Data Mining (RDM). The specific MMRF application to annotate human protein with diseases verifies that group knowledge (mainly protein-protein interaction pairs) improves the prediction, particularly doubling the area under the precision-recall curve
La aplicación de modelos de consciencia artificial en los sistemas multiagente
(Universidad de Castilla-La Mancha. Departamento de Sistemas Informáticos, 2006) Arrabales, Raúl; Sanchis de Miguel, María Araceli
Durante la última década han aparecido algunas implementaciones de modelos científicos de la consciencia basados en sistemas multiagente. El propósito de este artículo es recopilar y describir estos sistemas, determinando hasta que punto estas implementaciones satisfacen los modelos correspondientes, y analizando si proporcionan realmente las supuestas ventajas de usar consciencia artificial en la resolución de problemas. También se analizan en general las funciones de la consciencia y los beneficios que éstas pueden aportar en el rendimiento de los sistemas multiagente.
Comparing behavior in agent modelling task
(International Association for Development of the Information Society (IADIS), 2006) Iglesias Martínez, José Antonio; Ledezma Espino, Agapito Ismael; Sanchis de Miguel, María Araceli
In multi-agent system, agents have to analyze several features in order to adapt their behavior to the current situation. This extracted information is usually related to the environment and other agents influence. In this paper we present a method that compare two different agent models in order to extract the qualitative differences between them. This proposed comparative method captures several features of the two agent models and model them considering its behavior.
Sequence classification using statistical pattern recognition
(Springer, 2007) Iglesias Martínez, José Antonio; Ledezma Espino, Agapito Ismael; Sanchis de Miguel, María Araceli
Sequence classification is a significant problem that arises in many different real-world applications. The purpose of a sequence classifier is to assign a class label to a given sequence. Also, to obtain the pattern that characterizes the sequence is usually very useful. In this paper, a technique to discover a pattern from a given sequence is presented followed by a general novel method to classify the sequence. This method considers mainly the dependencies among the neighbouring elements of a sequence. In order to evaluate this method, a UNIX command environment is presented, but the method is general enough to be applied to other environments.
Verifying RoboCup Teams
(Springer, 2009) Benac Earle, Clara; Fredlund, Lars-Ake; Iglesias Martínez, José Antonio; Ledezma Espino, Agapito Ismael
Verification of multi-agent systems is a challenging task due to their dynamic nature, and the complex interactions between agents. An example of such a system is the RoboCup Soccer Simulator, where two teams of eleven independent agents play a game of football against each other. In the present article we attempt to verify a number of properties of RoboCup football teams, using a methodology involving testing. To accomplish such testing in an efficient manner we use the McErlang model checker, as it affords precise control of the scheduling of the agents, and provides convenient access to the internal states and actions of the agents of the football teams.
A comparing method of two team behaviours in the simulation coach competition
(Springer, 2006) Iglesias Martínez, José Antonio; Ledezma Espino, Agapito Ismael; Sanchis de Miguel, María Araceli
The main goal of agent modelling is to extract and represent the knowledge about the behaviour of other agents. Nowadays, modelling an agent in multi-agent systems is increasingly becoming more complex and significant. Also, robotic soccer domain is an interesting environment where agent modelling can be used. In this paper, we present an approach to classify and compare the behaviour of a multi-agent system using a coach in the soccer simulation domain of the RoboCup.
An efficient behavior classifier based on distributions of relevant events
(IOS Press, 2008) Iglesias Martínez, José Antonio; Ledezma Espino, Agapito Ismael; Sanchis de Miguel, María Araceli; Kaminka, Gal
Classifying efficiently the behavior of a Soccer team
(IOS Press, 2008) Iglesias Martínez, José Antonio; Ledezma Espino, Agapito Ismael; Sanchis de Miguel, María Araceli; Kaminka, Gal
In order to make a good decision, humans usually try to predict the behavior of others. By this prediction, many different tasks can be performed, such as to coordinate with them, to assist them or to predict their future behavior. In competitive domains, to recognize the behavior of the opponent can be very advantageous. In this paper, an approach for creating automatically the model of the behavior of a soccer team is presented. This approach is an effective and notable improvement of a previous work. As the actions performed by a soccer team are sequential, this sequentiality should be considered in the modeling process. Therefore, the observations of a soccer team in a dynamic, complex and continuous multi-variate world state are transformed into a sequence of atomic behaviors. Then, this sequence is analyzed in order to find out a model that defines the team behavior. Finally, the classification of an observed team is done by using a statistical test.
Modelling evolving user behaviours
(IEEE, 2009-05) Iglesias Martínez, José Antonio; Plamen, Angelov; Ledezma Espino, Agapito Ismael; Sanchis de Miguel, María Araceli
Knowledge about computer users is very beneficial for assisting them, predicting their future actions or detecting masqueraders. In this paper, a new approach for creating and recognizing automatically the behaviour profile of a computer user is presented. In this case, a computer user behaviour is represented as the sequence of the commands (s)he types during her/his work. This sequence is transformed into a distribution of relevant subsequences of commands in order to find out a profile that defines its behaviour. Also, because of a user profile is not necessarily fixed but rather it evolves/changes, we propose an evolving method to keep up to date the created profiles using an Evolving Systems approach. In this paper we combine the evolving classifier with a trie-based user profiling to obtain a powerful self-learning on-line scheme. We also develop further the recursive formula of the potential of a data point to become a cluster centre using cosine distance which is provided in the Appendix. The novel approach proposed in this paper can be applicable to any problem of dynamic/evolving user behaviour modelling where it can be represented as a sequence of actions and events. It has been evaluated on several real data streams.
Creating user profiles from a command-line interface: a statistical approach
(Springer, 2009) Iglesias Martínez, José Antonio; Ledezma Espino, Agapito Ismael; Sanchis de Miguel, María Araceli
Knowledge about computer users is very beneficial for assisting them, predicting their future actions or detecting masqueraders. In this paper, an approach for creating and recognizing automatically the behavior profile of a user from the commands (s)he types in a commandline interface, is presented. Specifically, in this research, a computer user behavior is represented as a sequence of UNIX commands. This sequence is transformed into a distribution of relevant subsequences in order to find out a profile that defines its behavior. Then, statistical methods are used for recognizing a user from the commands (s)he types. The experiment results, using 2 different sources of UNIX command data, show that a system based on our approach can efficiently recognize a UNIX user. In addition, a comparison with a HMM-base method is done. Because a user profile usually changes constantly, we also propose a method to keep up to date the created profiles using an age-based mechanism.
Human activity recognition in intelligent home environments: an evolving approach
(IOS Press, 2010) Iglesias Martínez, José Antonio; Angelov, Plamen; Ledezma Espino, Agapito Ismael; Sanchis de Miguel, María Araceli
In this paper, we propose an automated approach to track and recognize daily activities. Any activity is represented in this research as a sequence of raw sensors data. These sequences are treated using statistical methods in order to discover activity patterns. However, as the way to perform an activity is usually not fixed but it changes and evolves, we propose an activity recognition method based on Evolving Systems.
The winning advantage: using opponent models in robot Soccer
(Springer, 2009) Iglesias Martínez, José Antonio; Fernández, Juan Antonio; Villena, Ignacio Ramón; Ledezma Espino, Agapito Ismael; Sanchis de Miguel, María Araceli
Opponent modeling is a skill in multi-agent systems (MAS) which attempts to create a model of the behavior of the opponent. This model can be used to predict the future actions of the opponent and generate appropriate strategies to play against it. Several researches present different methods to create an opponent model in the RoboCup environment. However, how these models can impact the performance of teams is an essential aspect. This paper introduces a novel approach to use efficiently opponent models in order to improve our own team behavior. The basis of this approach is the research done by CAOS Coach Team for modeling and recognizing behaviors evaluated in the RoboCup Coach Competition 2006. For using these models, it is necessary a special agent (coach) which can model the observed opponent team (based on the previous research) and communicate a counter-strategy to the coached players (using the approach proposed in this paper). The evaluation of this approach is a hard problem, but we have conducted several experiments that can help us to know if we are going in a promising direction.

Browse

Recent Submissions