Emergent behaviour: learning from an artificially intelligent performing software agent

This research investigates the possibility for emergent choreographic behaviour to arise from the interactions between a human dancer and a learning, digital performing agent. The cognitive framework is extended through theories of distributed cognition to take into account the two interacting agents rather than a single agent and its environment. The Artificial Neural Network based performing agent demonstrated emergent dance behaviour when performing live with the human dancer. The agent was able to follow the dancer, create movement phrases based on what the dancer was performing and recognize short movement phrases, as a result of the interaction of the dancer’s motion captured movement data and the agent’s artificial neural network. This emergent behaviour was not explicitly programmed, but emerged as a result of the learning process and the interactions with the human dancer.

Responding to human full-body gestures embedded in motion data streams.

 This research created a neural-network enabled artificially intelligent performing agent that was able to learn to dance and recognise movement through a rehearsal and performance process with a human dancer. The agent exhibited emergent dance behaviour and successfully engaged in a live, semi-improvised dance performance with the human dancer.

Emergence

Emergence, by John McCormick, Steph Hutchison and an emerging performing agent, is a duet performed between a human dancer and an artificially intelligent performing agent. The work premiered on 1 August 2014 at Metanoia Theatre, Brunswick.

Duet: Improvising spatial dialogues with an artificially intelligent agent

This paper presents an experimental framework for a virtual reality artwork, Duet, that employs a combination of live, full body motion capture and Oculus Rift HMD to construct an experience through which a human User can spatially interact with an artificially intelligent Agent. The project explores conceptual notions of embodied knowledge transfer, shared poetics of movement and distortions of the body schema. Within this context, both the User and the Agent become performers, constructing an intimate and spontaneously generated proximal space. The project generates a visualization of the relationship between the User and the Agent without the context of a fixed VR landscape or architecture. The Agent's ability to retain and accumulate movement knowledge in a way that mimics human learning transforms an interactive experience into a collaborative one. The virtual representation of both performers is distorted and amplified in a dynamic manner, enhancing the potential for creative dialogue between the Agent and the User.

Teaching a digital performing agent: artificial neural network and hidden Markov Model for recognising and performing dance movement

For a Digital Performing Agent to be able to perform live with a human dancer, it would be useful for the agent to be able to contextualize the movement the dancer is performing and to have a suitable movement vocabulary with which to contribute to the performance. In this paper we will discuss our research into the use of Artificial Neural Networks (ANN) as a means of allowing a software agent to learn a shared vocabulary of movement from a dancer. The agent is able to use the learnt movements to form an internal representation of what the dancer is performing, allowing it to follow the dancer, generate movement sequences based on the dancer's current movement and dance independently of the dancer using a shared movement vocabulary. By combining the ANN with a Hidden Markov Model (HMM) the agent is able to recognize short full body movement phrases and respond when the dancer performs these phrases. We consider the relationship between the dancer and agent as a means of supporting the agent's learning and performance, rather than developing the agent's capability in a self-contained fashion.

Recognition: combining human interaction and a digital performing agent

Virtual and augmented environments are often dependent on human intervention for change to occur. However there are times when it would be advantageous for appropriate human-like activity to still occur when there are no humans present. In this paper, we describe the installation art piece Recognition, which uses the movement of human participants to effect change, and the movement of a performing agent when there are no humans present. The agent's Artificial Neural Network has learnt appropriate movements from a dancer and is able to generate suitable movement for the main avatar in the absence of human participants.

Collaborative dance between robot and human

Dance is an inherently embodied activity. The dancer is attuned to the effects of the physical world on her own physicality and the relationship of her presence to other dancers. This research is an investigation into artificially intelligent performing agents and robots and how a human dancer can guide the learning and performance of a robot performer. Using Artificial Neural Networks as the bases for the agent’s computational intelligence, performing agents were created that can perform by collaborating with human dancers through robots.

Building agent-based hybrid intelligent systems : a case study

Many complex problems (e.g., financial investment planning, foreign exchange trading, data mining from large/multiple databases) require hybrid intelligent systems that integrate many intelligent techniques (e.g., fuzzy logic, neural networks, and genetic algorithms). However, hybrid intelligent systems are difficult to develop because they have a large number of parts or components that have many interactions. On the other hand, agents offer a new and often more appropriate route to the development of complex systems, especially in open and dynamic environments. Thus, this paper discusses the development of an agent-based hybrid intelligent system for financial investment planning, in which a great number of heterogeneous computing techniques/packages are easily integrated into a unifying agent framework. This shows that agent technology can indeed facilitate the development of hybrid intelligent systems.

A multi-agent framework for financial investment planning

If a company or person wants to invest a lot of money, where, when, and how should the investment go? A multi-agent based Financial Investment Planner may give some reasonable answers to the above question. Good advice is mainly based on adequate information, rich knowledge, and great
skills to use knowledge and information. To this end, this planner consists of four principal components information gathering agents that are responsible for gathering relevant information on the Internet, data mining agents that are in charge of discovering knowledge from retrieved information as well as other relevant databases, group decision making agents that can effectively use available knowledge and appropriate information to make reasonable decisions (investment advice), and a graphical user interface that interacts with users. This paper is focused on the group decision making part. The design and implementation of an agent-based hybrid intelligent system - agent-based soft computing society are detailed.

Modelling financial investment planning from agent perspectives

Multi-agent Systems (MASs) offer strong models for representing complex and dynamic real-world environments. Taking financial investment planning as an example, this paper describes how to model complex systems from agent perspectives. Different agents and their behaviours are identified for financial investment planning. These agents are put together as an agent-based system. The experimental results show that all agents in the system can work cooperatively to provide reasonable investment advice. The system is very flexible and robust. The success of the system indicates that (MASs) can significantly facilitate the modelling of complex systems.

Beyond Simon’s means-ends analysis : natural creativity and the unanswered ‘why’ in the design of intelligent systems for problem-solving

Goal-directed problem solving as originally advocated by Herbert Simon’s means-ends analysis model has primarily shaped the course of design research on artificially intelligent systems for problem-solving. We contend that there is a definite disregard of a key phase within the overall design process that in fact logically precedes the actual problem solving phase. While systems designers have traditionally been obsessed with goal-directed problem solving, the basic determinants of the ultimate desired goal state still remain to be fully understood or categorically defined. We propose a rational framework built on a set of logically interconnected conjectures to specifically recognize this neglected phase in the overall design process of intelligent systems for practical problem-solving applications.

A multi-agent approach for enhancing transient stability of smart grids

Transient stability, an important issue to avoid the loss of synchronous operation in power systems, can be achieved through proper coordination and operation of protective devices within the critical clearing time (CCT). In view of this, the development of an intelligent decision support system is useful for providing better protection relay coordination. This paper presents an intelligent distributed agent-based scheme to enhance the transient stability of smart grids in light of CCT where a multi-agent framework (MAF) is developed and the agents are represented in such a way that they are equipped with protection relays (PRs). In addition to this, an algorithm is developed which assists the agents to make autonomous decision for controlling circuit breakers (CBs) independently. The proposed agents are responsible for the coordination of protection devices which is done through the precise detection and isolation of faults within the CCT. The agents also perform the duty of reclosing CBs after the clearance of faults. The performance of the proposed approach is demonstrated on a standard IEEE 39-bus test system by considering short-circuit faults at different locations under various load conditions. To further validate the suitability of the proposed scheme a benchmark 16-machine 68-bus power system is also considered. Simulation results show that MAF exhibits full flexibility to adapt the changes in system configurations and increase the stability margin for both test systems.

The Pinoke Project Performance

The Pinoke Project is a 6 week creative development to create a full-length transmedia dance performance at the cutting edge of artificially intelligent technology, elite contemporary dance practice, and publication. This process will be a collaboration between creative coders, 3D graphics and motion artists, a dance/choreographer, and an embedded dance critic who will work with an artificially intelligent robot, Pinoke, on the creative development of a new stage production as well as the documentation and dissemination of the process. This project is supported by the Victorian Government through Creative Victoria.

Agent-Based Hybrid Intelligent Systems: An Agent-Based Framework for Complex Problem Solving

Pt. I. Fundamentals of hybrid intelligent systems and agents -- 1. Introduction -- 2. Basics of hybrid intelligent systems -- 3. Basics of agents and multi-agent systems -- Pt. II. Methodology and framework -- 4. Agent-oriented methodologies -- 5. Agent-based framework for hybrid intelligent systems --6. Matchmaking in middle agents -- Pt. III. Application systems -- 7. Agent-based hybrid intelligent system for financial investment
planning -- 8. Agent-based hybrid intelligent system for data mining -- Pt. IV. Concluding remarks -- 9. The less the more -- App. Sample source codes of the agent-based financial planning system