Social presence, the sense of connection with another, is more important than ever as teachers, h... more Social presence, the sense of connection with another, is more important than ever as teachers, healthcare providers, and other professionals are using immersive tools to facilitate the social interaction for education, training, therapy and collaboration between geographically distributed humans and surrogates (avatars, agents, or robots). Leading researchers cite the subjective nature of the traditional self-report measures of social presence and the absence of a standardized approach to measuring social presence as a constraint to gaining deeper understanding of user’s experiences of emerging and existing tools. This discourse highlights behavioral indicators of social presence that have been identified over decades across disciplines from psychology, communication, computer science, education, and engineering. The authors explicate the behavioral themes of social presence and describe a classification system grounded in exogenic and endogenic themes of social presence. This arti...
This paper proposes a new computational model for parallel and distributed systems (PDSs). Tradit... more This paper proposes a new computational model for parallel and distributed systems (PDSs). Traditional sequential models of computation (e.g. Turing machines) do not possess the level of abstraction needed for reasoning about critical aspects of parallel and distributed computation. Augmenting a sequential system with additional processors introduces signi cant complexities associated with managing multiple concurrent activities this is just the parallel aspect of PDSs. Parallel processors that are also distributed across multiple remote locations introduce further complexities associated with the physical limitations of long distance communication. A computational model for PDSs must at once abstract away cumbersome details while providing a mechanism that promotes reasoning about important aspects of such systems. As a rst step toward achieving such a model, this paper presents an operational semantics for actors, a widely employed paradigm for PDS computation.
The development of distributed applications has not progressed as rapidly as its enabling technol... more The development of distributed applications has not progressed as rapidly as its enabling technologies. In part, this is due to the difficulty of reasoning about such complex systems. One reason for the added complexity is the need for communication within modern computing systems. In contrast to sequential systems, parallel systems give rise to parallel events (communications), and the resulting uncertainty of the observed order of these events. Loosely coupled distributed systems complicate this even further by introducing the element of multiple imperfect observers of these parallel events. To address these challenges, we introduce view-centric reasoning, an approach to thinking about modern computing systems that directly supports multiple, inconsistent and imperfect views of computation. While view-centric reasoning is general enough to apply to any communication model, the focus of this paper is on computing systems that employ generative communication, a middleware-based distributed shared memory manipulated by a coordination (communication) language. In particular, we apply view-centric reasoning to tuple space based systems and the Linda coordination language. View-centric reasoning helps us resolve a potential ambiguity in the semantics of Linda predicate operations found in commercial implementations of tuple space, such as Sun's JavaSpaces and IBM's T Spaces.
Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 2002
This research assessed the importance of rendering specific details when creating a virtual fores... more This research assessed the importance of rendering specific details when creating a virtual forest. Specifically, we examined memory for computer-generated trees using a modified recognition task in which participants were shown a target tree, engaged in a distractor task, and then ranked the similarity of seven foils to the original tree they had seen. Five of the foils represented changes on only one dimension of the tree whereas the other two foils represented modifications to either five features previously identified as salient or all nine tree features. Results showed that similarity rankings were largely based on overall structural similarity of the trees as opposed to similarity on smaller details such as branch thickness or leaf size. Additionally, perceived similarity rankings varied as a function of the symmetry of the tree. Virtual forests need to show realism for different features depending on the forest type.
Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 2001
Recent research on the design of virtual environments has focused on the important perceptual cha... more Recent research on the design of virtual environments has focused on the important perceptual characteristics of man-made "carpentered" environments, rather than on VEs of natural environments. The present research examines memory for characteristics of natural settings consisting of virtual trees. Participants viewed either a symmetrical or asymmetrical virtual tree and then recreated it using custom-designed tree editing software. Memory was more accurate for the symmetrical tree. Across trees, participants were most accurate recreating gross structural dimensions of a tree such as height and leaf size, and were particularly inaccurate at recreating the curvature of tree branches. Our conclusion is that the design of virtual environments should focus on accurately representing gross structural properties of trees, rather than on using high levels of detail to accurately portray trunk and branch curvature.
Proceedings of the 11th Annual conference on Genetic and evolutionary computation, 2009
A major goal for researchers in neuroevolution is to evolve artificial neural networks (ANNs) tha... more A major goal for researchers in neuroevolution is to evolve artificial neural networks (ANNs) that can learn during their lifetime. Such networks can adapt to changes in their environment that evolution on its own cannot anticipate. However, a profound problem with evolving adaptive systems is that if the impact of learning on the fitness of the agent is only marginal, then evolution is likely to produce individuals that do not exhibit the desired adaptive behavior. Instead, because it is easier at first to improve fitness without evolving the ability to learn, they are likely to exploit domaindependent static (i.e. non-adaptive) heuristics. This paper proposes a way to escape the deceptive trap of static policies based on the novelty search algorithm, which opens up a new avenue in the evolution of adaptive systems because it can exploit the behavioral difference between learning and nonlearning individuals. The main idea in novelty search is to abandon objective-based fitness and instead simply search only for novel behavior, which avoids deception entirely and has shown prior promising results in other domains. This paper shows that novelty search significantly outperforms fitness-based search in a tunably deceptive T-Maze navigation domain because it fosters the emergence of adaptive behavior.
A number of colleges and universities have recently added new degree programs in Information Tech... more A number of colleges and universities have recently added new degree programs in Information Technology (IT), or added IT components to existing Computer Science (CS) programs. Java language and technology are almost inescapable elements of both CS and IT programs. One of Java's more advanced features, languagelevel support for concurrency in the form of explicit multithreading, is important to both CS and IT students, but for different reasons. Teaching Java concurrency to CS and IT students, therefore, presents different challenges and requires emphasizing the topic in different ways. We discuss these issues, and present our experiences from CS and IT classes taught recently, in which Java concurrency was a topic.
The problems associated with Linda predicate operations inp() and rdp() are well known but not al... more The problems associated with Linda predicate operations inp() and rdp() are well known but not always well understood. One problem, from a purely academic standpoint, is that of semantic ambiguity in the case of failure. Despite this problem, commercial tuple space implementations all provide some version of the Linda predicate operations, and thus, problems concerning safety and liveness arise. Solutions to address these problems range from proposing extensions to an existing model of computation (CSP) to proposing alternative definitions for the Linda predicate operations themselves. While these solutions may at first appear unrelated, it is possible to relate these two disparate approaches in the context of lazy and eager semantics. The benefits of this characterization include a clearer understanding of the Linda predicates ambiguity problem, two respective solutions to this problem, and the importance of considering both lazy and eager semantic perspectives as techniques for problem solving in parallel and distributed systems.
Distributed computing systems, including those that uti- lize space-based middleware, present sig... more Distributed computing systems, including those that uti- lize space-based middleware, present significant challenges when attempting to reason formally about their behaviors and properties. In particular, two or more computational events may occur in parallel. We introduce View-Centric Reasoning (VCR)—a set of abstractions that comprises a general framework for reasoning about parallel and dis- tributed computation. First we extend the CSP metaphor to support traces of parallel events, rather than the tradi- tional random interleaving of individual events. Next we introduce the concept of views to represent explicitly the multiple possible perspectives of the same parallel compu- tation. Finally, we consider an instance of VCR for reason- ing about Gelernter's Linda language and tuple space com- putation, the basis for much of today's space-based middle- ware.
Journal of Visual Communication and Image Representation, 2007
A common method to create high dynamic range (HDR) images is to combine several different exposur... more A common method to create high dynamic range (HDR) images is to combine several different exposures of the same scene. In this approach, the use of higher ISO settings will reduce exposure times, and thereby the total capture time. This is advantageous in certain environments where it may help minimize ghosting artifacts. However, exposures taken at high sensitivity settings tend to be noisy, which is further amplified by the HDR creation algorithm. We present a robust and efficient technique to significantly reduce noise in an HDR image even when its constituent exposures are taken at very high ISO settings. The method does not introduce blur or other artifacts, and leverages the wealth of information available in a sequence of aligned exposures.
In contrast to sequential computation, concurrent computation gives rise to parallel events. Effo... more In contrast to sequential computation, concurrent computation gives rise to parallel events. Efforts to translate the history of concurrent computations into sequential event traces result in the potential uncertainty of the observed order of these events. Loosely coupled distributed systems complicate this uncertainty even further by introducing the element of multiple imperfect observers of these parallel events. Properties of such systems are difficult to reason about, and in some cases, attempts to prove safety or liveness lead to ambiguities. We present a survey of challenges of reasoning about properties of concurrent systems. We then propose a new approach, view-centric reasoning, that avoids the problem of translating concurrency into a sequential representation. Finally. we demonstrate the usefulness of view-centric reasoning as a framework for disambiguating the meaning of tuple space predicate operations, versions of which exist commercially in IBM's T Spaces and Sun's JavaSpaces.
Both Lawrence's HCSP [3] and Smith, et al's VCR [6] (an earlier version appears in [5]) extend CS... more Both Lawrence's HCSP [3] and Smith, et al's VCR [6] (an earlier version appears in [5]) extend CSP [1] with representations of truly concurrent events. Previously, VCR was described using an operational semantics, while the semantics of HCSP's Acceptances model, like those of the predominant CSP models described by Roscoe [4] (e.g., Traces, Failures / Divergences), are denotational. We now present a denotational semantics for VCR and, in so doing, propose an extension to HCSP (and possibly other existing CSP models) to support View-Centric Reasoning. This work brings VCR a step closer to being drawn within Hoare and He's Unifying Theories of Programming [2] for further comparisons.
Annals of Mathematics and Artificial Intelligence, 1996
CLP(~) is a constraint logic programming language in which constraints can be expressed in the do... more CLP(~) is a constraint logic programming language in which constraints can be expressed in the domain of real numbers. Computation in this specialized domain gives access to information useful in intelligent backtracking. In this paper, we present an efficient constraint satisfaction algorithm for linear constraints in the real number domain and show that our algorithm directly generates minimal sets of conflicting constraints when failures occur. We demonstrate how information gleaned during constraint satisfaction can be integrated with unification failure analysis. The resulting intelligent backtracking method works in the context of a two-sorted domain, where variables can be bound to either structured terms or real number expressions. We discuss the implementation of backtracking and show examples where the benefit of pruning the search tree outweighs the overhead of failure analysis.
This article presents in an informal way some early results on the design of a series of paradigm... more This article presents in an informal way some early results on the design of a series of paradigms for visualization of the parallel execution of logic programs. The results presented here refer to the visualization of or-parallelism, as in MUSE and Aurora, deterministic dependent and-parallelism, as in Andorra-I, and independent and-parallelism as in &-Prolog. A tool has been implemented for this purpose and has been interfaced with these systems. Results are presented showing the visualization of executions from these systems and the usefulness of the resulting tool is briefly discussed. 'There are two different visualization tools with the same name: the one we are currently referring to and the ParaGraph by Aikawa et al., described in [AKK + 92], which we refer to below.
Modeling and rendering of natural scenes based on Lindenmayer-systems (L-systems) has been employ... more Modeling and rendering of natural scenes based on Lindenmayer-systems (L-systems) has been employed in graphical simulations for several decades. Its main use has been in generating photo-realistic images of plants. However, L-plants (plants that are generated using L-systems) have not been used in real-time environments due to their rendering complexity. The goal of our research is to develop and implement algorithms that dynamically generate Level-of-Detail (LOD) models for L-plants. The recursive coalescence of stems and leaf clusters used to generate LOD models is computed online based on user-specified LOD ranges and reduction percentages, as well as plant characteristics such as height and radius. The prototype implementation reported here simulates a complex, interactive scene consisting of Loaks and L-palmettos.
Biological brains can adapt and learn from past experience. Yet neuroevolution, that is, automati... more Biological brains can adapt and learn from past experience. Yet neuroevolution, that is, automatically creating artificial neural networks (ANNs) through evolutionary algorithms, has sometimes focused on static ANNs that cannot change their weights during their lifetime. A profound problem with evolving adaptive systems is that learning to learn is highly deceptive. Because it is easier at first to improve fitness without evolving the ability to learn, evolution is likely to exploit domain-dependent static (i.e., nonadaptive) heuristics. This article analyzes this inherent deceptiveness in a variety of different dynamic, reward-based learning tasks, and proposes a way to escape the deceptive trap of static policies based on the novelty search algorithm. The main idea in novelty search is to abandon objective-based fitness and instead simply search only for novel behavior, which avoids deception entirely. A series of experiments and an in-depth analysis show how behaviors that could ...
For individuals affected by Autism Spectrum Disorder (ASD), the inability to make eye contact is ... more For individuals affected by Autism Spectrum Disorder (ASD), the inability to make eye contact is a significant barrier to their engagement in social environments. This lack of eye contact limits their ability to read social and emotional cues exhibited through facial expressions resulting in a corresponding decrease in social engagement. The use of interactive virtual environments (VEs) as a therapeutic protocol is a growing field of study. In recent studies, individuals with ASD were placed in VEs and engaged with avatars controlled by a human in the background, resulting in improvements in eye contact and engagement for some subjects. This paper is the first in a series of experiments exploring the potential of virtual avatars controlled through software agency, rather than human control, as a therapeutic tool for ASD. This paper examines if a subject could learn to make eye contact with an avatar and consequently recognize and respond to emotional cues expressed by the avatar. Results indicate that children with ASD can learn to recognize the emotional cues of the virtual avatar, and that their reactions to the avatar's needs as well as their eye contact with the avatar improved over the course of the VE experiment. This study sets the stage for future exploration into therapeutic use of agent-based virtual avatars, including transference of emotional cues from avatars to humans in the real world.
The requirements for tracking in augmented reality environments are stringent because of the need... more The requirements for tracking in augmented reality environments are stringent because of the need to register real and computer-generated virtual objects. Driven by the need to track real objects within these environments, we propose two algorithms to distribute markers on complex rigid objects. The proposed algorithms employ an optimization technique with a spherical or cylindrical intermediary surface. The validity and effectiveness of the algorithms are tested heuristically by simulation.
Environmental policy a b s t r a c t We develop the concept of virtual experiments and consider t... more Environmental policy a b s t r a c t We develop the concept of virtual experiments and consider their application to environmental policy. A virtual experiment combines insights from virtual reality in computer science, naturalistic decision-making from psychology, and field experiments from economics. The environmental policy applications of interest to us are the valuation of wild fire management policies such as prescribed burn. The methodological objective of virtual experiments is to bridge the gap between the artefactual controls of laboratory experiments and the naturalistic domain of field experiments or direct field studies. This should provide tools for policy analysis that combine the inferential power of replicable experimental treatments with the natural ‘‘look and feel’ ’ of a field domain. We present data from an experiment comparing valuations elicited by virtual experiments to those elicited by instruments that have some of the characteristics of standard survey ins...
Proceedings of the International Conference on Computer Graphics Theory and Applications, 2012
In Mixed Reality scenarios, background replacement is a common way to immerse a user in a synthet... more In Mixed Reality scenarios, background replacement is a common way to immerse a user in a synthetic environment. Properly identifying the background pixels in an image or video is a difficult problem known as matting. Proper alpha mattes usually come from human guidance, special hardware setups, or color dependent algorithms. This is a consequence of the under-constrained nature of the per pixel alpha blending equation. While the field of natural image matting has made progress finding a least squares solution for an alpha matte, the generation of trimaps, indicating regions of known foreground and background pixels, normally requires human interaction or offline computation. We overcome these limitations by combining a low fidelity depth image that segments the original video signal with a real-time parallel natural image matting technique that favors objects with similar colors in the foreground and background. This allows us to provide real-time alpha mattes for Mixed Reality scenarios that take place in relatively controlled environments. As a consequence, while monochromatic backdrops (such as green screens or retro-reflective material) aid the algorithm's accuracy, they are not an explicit requirement.
Social presence, the sense of connection with another, is more important than ever as teachers, h... more Social presence, the sense of connection with another, is more important than ever as teachers, healthcare providers, and other professionals are using immersive tools to facilitate the social interaction for education, training, therapy and collaboration between geographically distributed humans and surrogates (avatars, agents, or robots). Leading researchers cite the subjective nature of the traditional self-report measures of social presence and the absence of a standardized approach to measuring social presence as a constraint to gaining deeper understanding of user’s experiences of emerging and existing tools. This discourse highlights behavioral indicators of social presence that have been identified over decades across disciplines from psychology, communication, computer science, education, and engineering. The authors explicate the behavioral themes of social presence and describe a classification system grounded in exogenic and endogenic themes of social presence. This arti...
This paper proposes a new computational model for parallel and distributed systems (PDSs). Tradit... more This paper proposes a new computational model for parallel and distributed systems (PDSs). Traditional sequential models of computation (e.g. Turing machines) do not possess the level of abstraction needed for reasoning about critical aspects of parallel and distributed computation. Augmenting a sequential system with additional processors introduces signi cant complexities associated with managing multiple concurrent activities this is just the parallel aspect of PDSs. Parallel processors that are also distributed across multiple remote locations introduce further complexities associated with the physical limitations of long distance communication. A computational model for PDSs must at once abstract away cumbersome details while providing a mechanism that promotes reasoning about important aspects of such systems. As a rst step toward achieving such a model, this paper presents an operational semantics for actors, a widely employed paradigm for PDS computation.
The development of distributed applications has not progressed as rapidly as its enabling technol... more The development of distributed applications has not progressed as rapidly as its enabling technologies. In part, this is due to the difficulty of reasoning about such complex systems. One reason for the added complexity is the need for communication within modern computing systems. In contrast to sequential systems, parallel systems give rise to parallel events (communications), and the resulting uncertainty of the observed order of these events. Loosely coupled distributed systems complicate this even further by introducing the element of multiple imperfect observers of these parallel events. To address these challenges, we introduce view-centric reasoning, an approach to thinking about modern computing systems that directly supports multiple, inconsistent and imperfect views of computation. While view-centric reasoning is general enough to apply to any communication model, the focus of this paper is on computing systems that employ generative communication, a middleware-based distributed shared memory manipulated by a coordination (communication) language. In particular, we apply view-centric reasoning to tuple space based systems and the Linda coordination language. View-centric reasoning helps us resolve a potential ambiguity in the semantics of Linda predicate operations found in commercial implementations of tuple space, such as Sun's JavaSpaces and IBM's T Spaces.
Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 2002
This research assessed the importance of rendering specific details when creating a virtual fores... more This research assessed the importance of rendering specific details when creating a virtual forest. Specifically, we examined memory for computer-generated trees using a modified recognition task in which participants were shown a target tree, engaged in a distractor task, and then ranked the similarity of seven foils to the original tree they had seen. Five of the foils represented changes on only one dimension of the tree whereas the other two foils represented modifications to either five features previously identified as salient or all nine tree features. Results showed that similarity rankings were largely based on overall structural similarity of the trees as opposed to similarity on smaller details such as branch thickness or leaf size. Additionally, perceived similarity rankings varied as a function of the symmetry of the tree. Virtual forests need to show realism for different features depending on the forest type.
Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 2001
Recent research on the design of virtual environments has focused on the important perceptual cha... more Recent research on the design of virtual environments has focused on the important perceptual characteristics of man-made "carpentered" environments, rather than on VEs of natural environments. The present research examines memory for characteristics of natural settings consisting of virtual trees. Participants viewed either a symmetrical or asymmetrical virtual tree and then recreated it using custom-designed tree editing software. Memory was more accurate for the symmetrical tree. Across trees, participants were most accurate recreating gross structural dimensions of a tree such as height and leaf size, and were particularly inaccurate at recreating the curvature of tree branches. Our conclusion is that the design of virtual environments should focus on accurately representing gross structural properties of trees, rather than on using high levels of detail to accurately portray trunk and branch curvature.
Proceedings of the 11th Annual conference on Genetic and evolutionary computation, 2009
A major goal for researchers in neuroevolution is to evolve artificial neural networks (ANNs) tha... more A major goal for researchers in neuroevolution is to evolve artificial neural networks (ANNs) that can learn during their lifetime. Such networks can adapt to changes in their environment that evolution on its own cannot anticipate. However, a profound problem with evolving adaptive systems is that if the impact of learning on the fitness of the agent is only marginal, then evolution is likely to produce individuals that do not exhibit the desired adaptive behavior. Instead, because it is easier at first to improve fitness without evolving the ability to learn, they are likely to exploit domaindependent static (i.e. non-adaptive) heuristics. This paper proposes a way to escape the deceptive trap of static policies based on the novelty search algorithm, which opens up a new avenue in the evolution of adaptive systems because it can exploit the behavioral difference between learning and nonlearning individuals. The main idea in novelty search is to abandon objective-based fitness and instead simply search only for novel behavior, which avoids deception entirely and has shown prior promising results in other domains. This paper shows that novelty search significantly outperforms fitness-based search in a tunably deceptive T-Maze navigation domain because it fosters the emergence of adaptive behavior.
A number of colleges and universities have recently added new degree programs in Information Tech... more A number of colleges and universities have recently added new degree programs in Information Technology (IT), or added IT components to existing Computer Science (CS) programs. Java language and technology are almost inescapable elements of both CS and IT programs. One of Java's more advanced features, languagelevel support for concurrency in the form of explicit multithreading, is important to both CS and IT students, but for different reasons. Teaching Java concurrency to CS and IT students, therefore, presents different challenges and requires emphasizing the topic in different ways. We discuss these issues, and present our experiences from CS and IT classes taught recently, in which Java concurrency was a topic.
The problems associated with Linda predicate operations inp() and rdp() are well known but not al... more The problems associated with Linda predicate operations inp() and rdp() are well known but not always well understood. One problem, from a purely academic standpoint, is that of semantic ambiguity in the case of failure. Despite this problem, commercial tuple space implementations all provide some version of the Linda predicate operations, and thus, problems concerning safety and liveness arise. Solutions to address these problems range from proposing extensions to an existing model of computation (CSP) to proposing alternative definitions for the Linda predicate operations themselves. While these solutions may at first appear unrelated, it is possible to relate these two disparate approaches in the context of lazy and eager semantics. The benefits of this characterization include a clearer understanding of the Linda predicates ambiguity problem, two respective solutions to this problem, and the importance of considering both lazy and eager semantic perspectives as techniques for problem solving in parallel and distributed systems.
Distributed computing systems, including those that uti- lize space-based middleware, present sig... more Distributed computing systems, including those that uti- lize space-based middleware, present significant challenges when attempting to reason formally about their behaviors and properties. In particular, two or more computational events may occur in parallel. We introduce View-Centric Reasoning (VCR)—a set of abstractions that comprises a general framework for reasoning about parallel and dis- tributed computation. First we extend the CSP metaphor to support traces of parallel events, rather than the tradi- tional random interleaving of individual events. Next we introduce the concept of views to represent explicitly the multiple possible perspectives of the same parallel compu- tation. Finally, we consider an instance of VCR for reason- ing about Gelernter's Linda language and tuple space com- putation, the basis for much of today's space-based middle- ware.
Journal of Visual Communication and Image Representation, 2007
A common method to create high dynamic range (HDR) images is to combine several different exposur... more A common method to create high dynamic range (HDR) images is to combine several different exposures of the same scene. In this approach, the use of higher ISO settings will reduce exposure times, and thereby the total capture time. This is advantageous in certain environments where it may help minimize ghosting artifacts. However, exposures taken at high sensitivity settings tend to be noisy, which is further amplified by the HDR creation algorithm. We present a robust and efficient technique to significantly reduce noise in an HDR image even when its constituent exposures are taken at very high ISO settings. The method does not introduce blur or other artifacts, and leverages the wealth of information available in a sequence of aligned exposures.
In contrast to sequential computation, concurrent computation gives rise to parallel events. Effo... more In contrast to sequential computation, concurrent computation gives rise to parallel events. Efforts to translate the history of concurrent computations into sequential event traces result in the potential uncertainty of the observed order of these events. Loosely coupled distributed systems complicate this uncertainty even further by introducing the element of multiple imperfect observers of these parallel events. Properties of such systems are difficult to reason about, and in some cases, attempts to prove safety or liveness lead to ambiguities. We present a survey of challenges of reasoning about properties of concurrent systems. We then propose a new approach, view-centric reasoning, that avoids the problem of translating concurrency into a sequential representation. Finally. we demonstrate the usefulness of view-centric reasoning as a framework for disambiguating the meaning of tuple space predicate operations, versions of which exist commercially in IBM's T Spaces and Sun's JavaSpaces.
Both Lawrence's HCSP [3] and Smith, et al's VCR [6] (an earlier version appears in [5]) extend CS... more Both Lawrence's HCSP [3] and Smith, et al's VCR [6] (an earlier version appears in [5]) extend CSP [1] with representations of truly concurrent events. Previously, VCR was described using an operational semantics, while the semantics of HCSP's Acceptances model, like those of the predominant CSP models described by Roscoe [4] (e.g., Traces, Failures / Divergences), are denotational. We now present a denotational semantics for VCR and, in so doing, propose an extension to HCSP (and possibly other existing CSP models) to support View-Centric Reasoning. This work brings VCR a step closer to being drawn within Hoare and He's Unifying Theories of Programming [2] for further comparisons.
Annals of Mathematics and Artificial Intelligence, 1996
CLP(~) is a constraint logic programming language in which constraints can be expressed in the do... more CLP(~) is a constraint logic programming language in which constraints can be expressed in the domain of real numbers. Computation in this specialized domain gives access to information useful in intelligent backtracking. In this paper, we present an efficient constraint satisfaction algorithm for linear constraints in the real number domain and show that our algorithm directly generates minimal sets of conflicting constraints when failures occur. We demonstrate how information gleaned during constraint satisfaction can be integrated with unification failure analysis. The resulting intelligent backtracking method works in the context of a two-sorted domain, where variables can be bound to either structured terms or real number expressions. We discuss the implementation of backtracking and show examples where the benefit of pruning the search tree outweighs the overhead of failure analysis.
This article presents in an informal way some early results on the design of a series of paradigm... more This article presents in an informal way some early results on the design of a series of paradigms for visualization of the parallel execution of logic programs. The results presented here refer to the visualization of or-parallelism, as in MUSE and Aurora, deterministic dependent and-parallelism, as in Andorra-I, and independent and-parallelism as in &-Prolog. A tool has been implemented for this purpose and has been interfaced with these systems. Results are presented showing the visualization of executions from these systems and the usefulness of the resulting tool is briefly discussed. 'There are two different visualization tools with the same name: the one we are currently referring to and the ParaGraph by Aikawa et al., described in [AKK + 92], which we refer to below.
Modeling and rendering of natural scenes based on Lindenmayer-systems (L-systems) has been employ... more Modeling and rendering of natural scenes based on Lindenmayer-systems (L-systems) has been employed in graphical simulations for several decades. Its main use has been in generating photo-realistic images of plants. However, L-plants (plants that are generated using L-systems) have not been used in real-time environments due to their rendering complexity. The goal of our research is to develop and implement algorithms that dynamically generate Level-of-Detail (LOD) models for L-plants. The recursive coalescence of stems and leaf clusters used to generate LOD models is computed online based on user-specified LOD ranges and reduction percentages, as well as plant characteristics such as height and radius. The prototype implementation reported here simulates a complex, interactive scene consisting of Loaks and L-palmettos.
Biological brains can adapt and learn from past experience. Yet neuroevolution, that is, automati... more Biological brains can adapt and learn from past experience. Yet neuroevolution, that is, automatically creating artificial neural networks (ANNs) through evolutionary algorithms, has sometimes focused on static ANNs that cannot change their weights during their lifetime. A profound problem with evolving adaptive systems is that learning to learn is highly deceptive. Because it is easier at first to improve fitness without evolving the ability to learn, evolution is likely to exploit domain-dependent static (i.e., nonadaptive) heuristics. This article analyzes this inherent deceptiveness in a variety of different dynamic, reward-based learning tasks, and proposes a way to escape the deceptive trap of static policies based on the novelty search algorithm. The main idea in novelty search is to abandon objective-based fitness and instead simply search only for novel behavior, which avoids deception entirely. A series of experiments and an in-depth analysis show how behaviors that could ...
For individuals affected by Autism Spectrum Disorder (ASD), the inability to make eye contact is ... more For individuals affected by Autism Spectrum Disorder (ASD), the inability to make eye contact is a significant barrier to their engagement in social environments. This lack of eye contact limits their ability to read social and emotional cues exhibited through facial expressions resulting in a corresponding decrease in social engagement. The use of interactive virtual environments (VEs) as a therapeutic protocol is a growing field of study. In recent studies, individuals with ASD were placed in VEs and engaged with avatars controlled by a human in the background, resulting in improvements in eye contact and engagement for some subjects. This paper is the first in a series of experiments exploring the potential of virtual avatars controlled through software agency, rather than human control, as a therapeutic tool for ASD. This paper examines if a subject could learn to make eye contact with an avatar and consequently recognize and respond to emotional cues expressed by the avatar. Results indicate that children with ASD can learn to recognize the emotional cues of the virtual avatar, and that their reactions to the avatar's needs as well as their eye contact with the avatar improved over the course of the VE experiment. This study sets the stage for future exploration into therapeutic use of agent-based virtual avatars, including transference of emotional cues from avatars to humans in the real world.
The requirements for tracking in augmented reality environments are stringent because of the need... more The requirements for tracking in augmented reality environments are stringent because of the need to register real and computer-generated virtual objects. Driven by the need to track real objects within these environments, we propose two algorithms to distribute markers on complex rigid objects. The proposed algorithms employ an optimization technique with a spherical or cylindrical intermediary surface. The validity and effectiveness of the algorithms are tested heuristically by simulation.
Environmental policy a b s t r a c t We develop the concept of virtual experiments and consider t... more Environmental policy a b s t r a c t We develop the concept of virtual experiments and consider their application to environmental policy. A virtual experiment combines insights from virtual reality in computer science, naturalistic decision-making from psychology, and field experiments from economics. The environmental policy applications of interest to us are the valuation of wild fire management policies such as prescribed burn. The methodological objective of virtual experiments is to bridge the gap between the artefactual controls of laboratory experiments and the naturalistic domain of field experiments or direct field studies. This should provide tools for policy analysis that combine the inferential power of replicable experimental treatments with the natural ‘‘look and feel’ ’ of a field domain. We present data from an experiment comparing valuations elicited by virtual experiments to those elicited by instruments that have some of the characteristics of standard survey ins...
Proceedings of the International Conference on Computer Graphics Theory and Applications, 2012
In Mixed Reality scenarios, background replacement is a common way to immerse a user in a synthet... more In Mixed Reality scenarios, background replacement is a common way to immerse a user in a synthetic environment. Properly identifying the background pixels in an image or video is a difficult problem known as matting. Proper alpha mattes usually come from human guidance, special hardware setups, or color dependent algorithms. This is a consequence of the under-constrained nature of the per pixel alpha blending equation. While the field of natural image matting has made progress finding a least squares solution for an alpha matte, the generation of trimaps, indicating regions of known foreground and background pixels, normally requires human interaction or offline computation. We overcome these limitations by combining a low fidelity depth image that segments the original video signal with a real-time parallel natural image matting technique that favors objects with similar colors in the foreground and background. This allows us to provide real-time alpha mattes for Mixed Reality scenarios that take place in relatively controlled environments. As a consequence, while monochromatic backdrops (such as green screens or retro-reflective material) aid the algorithm's accuracy, they are not an explicit requirement.
Uploads
Papers by Charlie Hughes