scholarly journals Representing Qualitative  Action Models for Learning  in Complex Virtual Worlds

2021 ◽  
Author(s):  
◽  
Adam Clarke
Keyword(s):  
Virtual Worlds ◽  
Vision System ◽  
Expressive Power ◽  
Process Models ◽  
Benchmark Problems ◽  
Complex Environments ◽  
Advantages And Disadvantages ◽  
Finite State ◽  
Planning Algorithm ◽  
Description Languages

<p>This thesis addresses the problem of representing and learning qualitative models of behaviour in complex virtual worlds. It presents a novel representation, ‘Q-Systems’, that integrates two existing representation frameworks: qualitative process models and action description languages. QSystems combines the expressive power of both frameworks to allow actions and world dynamics to be modelled in a common way using a representation based on non-deterministic and probabilistic finite state machines. The representation supports learning and planning by using a modular approach that partitions world behaviour into ‘systems’ of objects with specific contexts and a related behaviour. Q-Systems was developed and tested using an agent in a rich simulated world that was created as part of the thesis. The simulation uses a rigid body physics engine to produce complex realistic interactions between objects. An action system and a qualitative vision system were also developed to allow the agent to observe and act in the simulated world. The thesis includes a proposed two stage learning process comprising an initial stage in which ‘histories’ (contextually and temporally restricted sequences of observations) are extracted from interactions with the simulation, and a second stage in which the histories are generalised to create a knowledge base of system models. An algorithm for generating histories is presented and a number of heuristics are implemented and compared. A system for learning generalised models is presented and it is used to assess the suitability of Q-Systems with respect to learning in complex environments. Planning with Q-Systems is demonstrated in an agent which reasons with generalised models to work out how to achieve goals in the simulated world. A simple planning algorithm is described and a variety of issues are explored. Planning with a single system is shown to be relatively straightforward due to the modular nature of Q-Systems. This thesis demonstrates that Q-Systems successfully integrate two different representation frameworks and that they can be used in learning and planning in complex environments. The initial results are promising, but further investigation is required to fully understand the advantages and disadvantages of the Q-System approach compared with existing learning systems. This would involve the development of benchmark problems (currently there are none for this particular domain).</p>

scholarly journals Representing Qualitative  Action Models for Learning  in Complex Virtual Worlds

2021 ◽  
Author(s):  
◽  
Adam Clarke
Keyword(s):  
Virtual Worlds ◽  
Vision System ◽  
Expressive Power ◽  
Process Models ◽  
Benchmark Problems ◽  
Complex Environments ◽  
Advantages And Disadvantages ◽  
Finite State ◽  
Planning Algorithm ◽  
Description Languages

<p>This thesis addresses the problem of representing and learning qualitative models of behaviour in complex virtual worlds. It presents a novel representation, ‘Q-Systems’, that integrates two existing representation frameworks: qualitative process models and action description languages. QSystems combines the expressive power of both frameworks to allow actions and world dynamics to be modelled in a common way using a representation based on non-deterministic and probabilistic finite state machines. The representation supports learning and planning by using a modular approach that partitions world behaviour into ‘systems’ of objects with specific contexts and a related behaviour. Q-Systems was developed and tested using an agent in a rich simulated world that was created as part of the thesis. The simulation uses a rigid body physics engine to produce complex realistic interactions between objects. An action system and a qualitative vision system were also developed to allow the agent to observe and act in the simulated world. The thesis includes a proposed two stage learning process comprising an initial stage in which ‘histories’ (contextually and temporally restricted sequences of observations) are extracted from interactions with the simulation, and a second stage in which the histories are generalised to create a knowledge base of system models. An algorithm for generating histories is presented and a number of heuristics are implemented and compared. A system for learning generalised models is presented and it is used to assess the suitability of Q-Systems with respect to learning in complex environments. Planning with Q-Systems is demonstrated in an agent which reasons with generalised models to work out how to achieve goals in the simulated world. A simple planning algorithm is described and a variety of issues are explored. Planning with a single system is shown to be relatively straightforward due to the modular nature of Q-Systems. This thesis demonstrates that Q-Systems successfully integrate two different representation frameworks and that they can be used in learning and planning in complex environments. The initial results are promising, but further investigation is required to fully understand the advantages and disadvantages of the Q-System approach compared with existing learning systems. This would involve the development of benchmark problems (currently there are none for this particular domain).</p>

SVM-based global vision system of sports competition and action recognition

2020 ◽  
pp. 1-12
Author(s):  
Changxin Sun ◽  
Di Ma
Keyword(s):  
Target Recognition ◽  
Vision System ◽  
Task Assignment ◽  
Recognition Algorithm ◽  
Control Factor ◽  
Training Time ◽  
Adaboost Algorithm ◽  
Advantages And Disadvantages ◽  
Time Control ◽  
Correlation Degree

In the research of intelligent sports vision systems, the stability and accuracy of vision system target recognition, the reasonable effectiveness of task assignment, and the advantages and disadvantages of path planning are the key factors for the vision system to successfully perform tasks. Aiming at the problem of target recognition errors caused by uneven brightness and mutations in sports competition, a dynamic template mechanism is proposed. In the target recognition algorithm, the correlation degree of data feature changes is fully considered, and the time control factor is introduced when using SVM for classification,At the same time, this study uses an unsupervised clustering method to design a classification strategy to achieve rapid target discrimination when the environmental brightness changes, which improves the accuracy of recognition. In addition, the Adaboost algorithm is selected as the machine learning method, and the algorithm is optimized from the aspects of fast feature selection and double threshold decision, which effectively improves the training time of the classifier. Finally, for complex human poses and partially occluded human targets, this paper proposes to express the entire human body through multiple parts. The experimental results show that this method can be used to detect sports players with multiple poses and partial occlusions in complex backgrounds and provides an effective technical means for detecting sports competition action characteristics in complex backgrounds.

scholarly journals A numerical study of fish adaption behaviors in complex environments with a deep reinforcement learning and immersed boundary–lattice Boltzmann method

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Yi Zhu ◽  
Fang-Bao Tian ◽  
John Young ◽  
James C. Liao ◽  
Joseph C. S. Lai
Keyword(s):  
Lattice Boltzmann Method ◽  
Lattice Boltzmann ◽  
Prey Capture ◽  
Numerical Study ◽  
Benchmark Problems ◽  
Immersed Boundary ◽  
Complex Environments ◽  
Fish Model ◽  
Point To Point ◽  
Boltzmann Method

AbstractFish adaption behaviors in complex environments are of great importance in improving the performance of underwater vehicles. This work presents a numerical study of the adaption behaviors of self-propelled fish in complex environments by developing a numerical framework of deep learning and immersed boundary–lattice Boltzmann method (IB–LBM). In this framework, the fish swimming in a viscous incompressible flow is simulated with an IB–LBM which is validated by conducting two benchmark problems including a uniform flow over a stationary cylinder and a self-propelled anguilliform swimming in a quiescent flow. Furthermore, a deep recurrent Q-network (DRQN) is incorporated with the IB–LBM to train the fish model to adapt its motion to optimally achieve a specific task, such as prey capture, rheotaxis and Kármán gaiting. Compared to existing learning models for fish, this work incorporates the fish position, velocity and acceleration into the state space in the DRQN; and it considers the amplitude and frequency action spaces as well as the historical effects. This framework makes use of the high computational efficiency of the IB–LBM which is of crucial importance for the effective coupling with learning algorithms. Applications of the proposed numerical framework in point-to-point swimming in quiescent flow and position holding both in a uniform stream and a Kármán vortex street demonstrate the strategies used to adapt to different situations.

Application of Combined Feedforward and Feedback Controller With Shaped Input to Benchmark Problem

2010 ◽  
Vol 132 (2) ◽  
Author(s):  
Young Joo Shin ◽  
Peter H. Meckl
Keyword(s):  
Feedback Control ◽  
Control Strategy ◽  
Control Strategies ◽  
Control Design ◽  
Single Frequency ◽  
Benchmark Problems ◽  
Control Force ◽  
Benchmark Problem ◽  
Advantages And Disadvantages ◽  
Robust Control Design

Benchmark problems have been used to evaluate the performance of a variety of robust control design methodologies by many control engineers over the past 2 decades. A benchmark is a simple but meaningful problem to highlight the advantages and disadvantages of different control strategies. This paper verifies the performance of a new control strategy, which is called combined feedforward and feedback control with shaped input (CFFS), through a benchmark problem applied to a two-mass-spring system. CFFS, which consists of feedback and feedforward controllers and shaped input, can achieve high performance with a simple controller design. This control strategy has several unique characteristics. First, the shaped input is designed to extract energy from the flexible modes, which means that a simpler feedback control design based on a rigid-body model can be used. In addition, only a single frequency must be attenuated to reduce residual vibration of both masses. Second, only the dynamics between control force and the first mass need to be considered in designing both feedback and feedforward controllers. The proposed control strategy is applied to a benchmark problem and its performance is compared with that obtained using two alternative control strategies.

scholarly journals Second Life as a research environment: avatar-based focus groups (AFG)

2016 ◽  
Vol 19 (1) ◽  
pp. 101-114 ◽  
Author(s):  
Eman Gadalla ◽  
Ibrahim Abosag ◽  
Kathy Keeling
Keyword(s):  
Focus Groups ◽  
Data Quality ◽  
Virtual Worlds ◽  
Second Life ◽  
Three Dimensional ◽  
Marketing Research ◽  
Two Phase ◽  
Content Type ◽  
Face To Face ◽  
Advantages And Disadvantages

Purpose – This study aims to examine the nature and the potential use of avatar-based focus groups (AFGs) (i.e. focus groups conducted in three-dimensional [3D] virtual worlds [VWs]) as compared to face-to-face and online focus groups (OFGs), motivated by the ability of VWs to stimulate the realism of physical places. Over the past decade, there has been a rapid increase in using 3D VWs as a research tool. Design/methodology/approach – Using a two-phase reflective approach, data were collected first by using traditional face-to-face focus groups, followed by AFGs. In Phase 2, an online, semi-structured survey provided comparison data and experiences in AFGs, two-dimensional OFGs and traditional face-to-face focus groups. Findings – The findings identify the advantages and disadvantages of AFGs for marketing research. There is no evident difference in data quality between the results of AFGs and face-to-face focus groups. AFG compensates for some of the serious limitations associated with OFGs. Practical implications – The paper reflects on three issues, data quality, conduct of AFGs (including the moderator reflection) and participant experience, that together inform one’s understanding of the characteristics, advantages and limitations of AFG. Originality/value – This is the first paper to compare between AFGs, traditional face-to-face focus groups and OFGs. AFG holds many advantages over OFGs and even, sometimes, over face-to-face focus groups, providing a suitable environment for researchers to collect data.

scholarly journals Copyful Streaming String Transducers

2021 ◽  
Vol 178 (1-2) ◽  
pp. 59-76
Author(s):  
Emmanuel Filiot ◽  
Pierre-Alain Reynier
Keyword(s):  
Polynomial Time ◽  
Equivalence Problem ◽  
Expressive Power ◽  
The Other ◽  
Finite State Automata ◽  
Finite State ◽  
Variable Content ◽  
Deterministic Automata ◽  
Intermediate Output ◽  
Linear Manner

Copyless streaming string transducers (copyless SST) have been introduced by R. Alur and P. Černý in 2010 as a one-way deterministic automata model to define transductions of finite strings. Copyless SST extend deterministic finite state automata with a set of variables in which to store intermediate output strings, and those variables can be combined and updated all along the run, in a linear manner, i.e., no variable content can be copied on transitions. It is known that copyless SST capture exactly the class of MSO-definable string-to-string transductions, and are as expressive as deterministic two-way transducers. They enjoy good algorithmic properties. Most notably, they have decidable equivalence problem (in PSpace). On the other hand, HDT0L systems have been introduced for a while, the most prominent result being the decidability of the equivalence problem. In this paper, we propose a semantics of HDT0L systems in terms of transductions, and use it to study the class of deterministic copyful SST. Our contributions are as follows: (i)HDT0L systems and total deterministic copyful SST have the same expressive power, (ii)the equivalence problem for deterministic copyful SST and the equivalence problem for HDT0L systems are inter-reducible, in quadratic time. As a consequence, equivalence of deterministic SST is decidable, (iii)the functionality of non-deterministic copyful SST is decidable, (iv)determining whether a non-deterministic copyful SST can be transformed into an equivalent non-deterministic copyless SST is decidable in polynomial time.

Motion Planning of an Innovative Serial Robot for Energy Consumption Reduction With “On Fly” Pieces Grasping

2008 ◽  
Author(s):  
Andrea Menegolo ◽  
Roberto Bussola ◽  
Diego Tosi
Keyword(s):  
Energy Consumption ◽  
Motion Planning ◽  
Cycle Time ◽  
Moving Objects ◽  
Vision System ◽  
Rotation Velocity ◽  
Simulation Software ◽  
Serial Robot ◽  
Planning Algorithm ◽  
Point To Point

The following study deals with the on-line motion planning of an innovative SCARA like robot with unlimited joint rotations. The application field is the robotic interception of moving objects randomly distributed on a conveyor and detected by a vision system. A motion planning algorithm was developed in order to achieve a satisfactory cycle time and energy consumption. The algorithm is based on the evaluation of the inertial actions arisen in the robot structure during the pick and place motions and it aims to keep constant the rotation velocity of the first joint during the motion, the grasping and the discarding phases. Since the algorithm must be applied run time and the number of the reachable pieces can be high, a particular care was dedicated to the computational burden reduction. Subsequently to an analytic study of the kinematical constraints and the criteria definition for the choice of which piece to grasp, a devoted simulation software was developed. The software allows the control and the evaluation of the effects of all the main parameters on the system behavior and a comparison of the cycle time and the energy consumption between the proposed algorithm and a standard point-to-point motion strategy.

Modelling blockchain-based business processes: a comparative analysis of BPMN vs CMMN

2021 ◽  
Vol 27 (2) ◽  
pp. 638-657
Author(s):  
Fredrik Milani ◽  
Luciano Garcia-Banuelos ◽  
Svitlana Filipova ◽  
Mariia Markovska
Keyword(s):  
Design Methodology ◽  
Business Processes ◽  
Process Analysis ◽  
Process Models ◽  
Disruptive Technology ◽  
Content Type ◽  
Advantages And Disadvantages ◽  
Blockchain Technology ◽  
Block Chain ◽  
Insight Into

PurposeBlockchain technology is increasingly positioned as a promising and disruptive technology. Such a promise has attracted companies to explore how blockchain technology can be used to gain significant benefits. Process models play a cardinal role when seeking to improve business processes as they are the foundation of process analysis and redesign. This paper examines how blockchain-oriented processes can be conceptually modelled with activity- (BPMN) and artifact-centric (CMMN) modelling paradigms.Design/methodology/approachThis paper discusses how commonly occurring patterns, specific to block-chain-based applications, can be modelled with BPMN and CMMN. Furthermore, the advantages and disadvantages of both notations for accurately representing blockchain-specific patterns are discussed.FindingsThe main finding of this paper is that neither BPMN nor CMMN can adequately and accurately represent certain patterns specific for blockchain-oriented processes. BPMN, while supporting most of the patterns, does not provide sufficient support to represent tokenization. CMMN, on the other hand, does not provide support to distinguish between activities executed and data stored on-chain versus off-chain.Originality/valueThe paper provides insight into the strengths and weaknesses of BPMN and CMMN for modelling processes to be supported by blockchain. This will serve to aid analysts to produce better process models for communication purposes and, thereby, facilitate development of blockchain-based solutions.

Digital Citizenship in Participatory Culture

2017 ◽  
pp. 60-73 ◽  
Author(s):  
Valerie J Hill
Keyword(s):  
Learning Environments ◽  
Information Literacy ◽  
Virtual Worlds ◽  
Personal Responsibility ◽  
Participatory Culture ◽  
Digital Culture ◽  
Digital Citizenship ◽  
Presentation Of Self ◽  
Advantages And Disadvantages ◽  
Collaborative Learning Opportunities

Global participatory digital culture provides collaborative learning opportunities beyond physical walls and without time constraints. Learners connect across the planet in real time. The virtual representation of self requires understanding the personal responsibility for digital citizenship and information literacy. Both the presentation of self and evaluation of content in all formats are new challenges for learners of all ages, including the youngest students born into an age of sharing and connecting. Virtual learning environments may transform education and certainly provide both advantages and disadvantages for educators and learners. Understanding the personal responsibility for digital citizenship is imperative to identify the best practices of education in virtual spaces. This chapter focuses on digital citizenship and information literacy in virtual worlds, virtual reality, and immersive learning environments.

The Role of Standards for E-Commerce in Virtual Worlds

2011 ◽  
pp. 257-266 ◽  
Author(s):  
Joerg H. Kloss
Keyword(s):  
Virtual Worlds ◽  
3D Graphics ◽  
Iso Standard ◽  
Advantages And Disadvantages ◽  
Future Developments ◽  
3D Environments ◽  
Interactive 3D ◽  
The Right

This chapter discusses the topic of standards for Virtual Worlds with emphasis on their usability as a stable and reliable basis for long-term investments into 3D-E-Commerce. The text explains why standards are important for the success of Virtual Worlds as well as the business in these shared online 3D environments, and what the relevant criteria are to decide for the right technology and/or provider. Although sometimes in the shadow of popular proprietary platforms there are already many different candidates for a Virtual World standard, currently in different states of development. By choosing a 3D platform, E-Commerce providers will decide about their business potential and at the same time strengthen one or another standard in the current technical competition phase. So it is important to get an overview about the current approaches, their advantages and disadvantages as well as the tendencies for the future developments. In this chapter the reader will be sensitized for the issues of standardization, compatibility and interoperability of Virtual Worlds for successful E-Commerce applications. An overview about the current approaches supports the orientation and decision for the different technologies. Some concrete XML-based code examples realized in the international ISO standard for interactive 3D-Graphics X3D demonstrates the practical deployment of highly compatible concepts. An outlook to the further integration of interactive 3D graphics into the Next Generation Web respectively the 3D Internet completes the overview.

Sign in / Sign up

Export Citation Format

Share Document