A Speech Interface to Virtual Environment

A Speech Interface to Virtual Environment

Authors

Scott McGlashan and Tomas Axling

Swedish Institute of Computer Science

Presentation Agenda Introduction The TALKING AGENT system

DIVE SR/TTS Agent Modeling Framework Interaction Metaphor Reference Resolution

Future Work Conclusion

Purposes of this paper Analyze the technical and design issues to combine a

virtual world with a speech interface. Describe system architecture of the TALKING

AGENT system.

Problems of Integration Speech Recognition : Limited vocabulary to gain

accuracy. Language Understanding : Limited knowledge to

maximize the understanding. Interaction Metaphor : Who does the user talk to?

(Above questions are discussed in detail in the authors’ last paper “Speech Interface to Virtual Reality”.)

Innovation of this System Combining intelligent agent and speech interface

to carry out specialized functions in the VR World. Functions have been implemented :

Transporting objects Fetching objects Painting objects Increasing the size of objects

System Architecture

DIVE-Virtual Reality System DIVE(Distribute Interactive Virtual Environment)

is a multi-user virtual environment. DIVE allow users and environment interact in real-

time. DIVE contains a database composed of

hierarchically organized objects .

DIME- DIVE Meeting Environment

Speech Recognition SR with limited pre-defined phrases promises good

recognition performance. Using grammar to set constraint to search space. Using commercial SR-engine (Nuance).

Agent Modeling Framework High-level languages do not support complex

symbolic computations. Oz is well suited for this purpose. Using ODI as interface between Oz and DIVE. The parent agent consists basic functions. We can define more specific agent by extend parent

agent.

Agent Modeling Framework

Interaction Metaphor Direct manipulation -Personal Presence. Various metaphors for spoken interaction have been

proposed. Proxy Divinity Telekinesis Interface Agent

This system adopt the Proxy metaphor.

The DIVERSE System-Interface Agent

Addressing Agent Inside the user’s eye-sight

Dialogue initiated by clicking on the agent.

Outside the user’s eye-sight Phone agent-First press the phone agent then connect to

remote agent

Feedback Given speech input ,system should give the visual

feedback to the user. If the agent listening or not?

What is the feedback when talking to agent far away?

Reference Resolution Given some descriptions , the reference resolution

engine maps them to object which user is referring to.

Considerations Object focus. Property Perception. Discourse Modeling.

Robust Interaction When errors don’t matter

User can view the results and current them by direct manipulation.

Safety-critical applications Confirm user command. Clarifying incomplete or ambiguous commands.

Future Work Agent behavior should related to its previous action . Add mental components. Talking to agent by aura-driven . Evaluate this system with realistic scenario.

Ex: virtual travel agency.

Conclusions Add a speech interface to VR-system. Using constraint SR to achieve high accuracy. Developing an appropriate metaphor. The agents modeled in this system provide specific

functions in the virtual world.

Q & A

Paper Source

McGlashan, S Speech Interfaces to Virtual Reality in Proceedings of the Second Conference on the Military Applications of Synthetic Environments and Virtual Reality, Stockholm, Sweden, 1995.