Hardware Requirements

For speech recognition:

- Pentium II PC 300Mhz or faster, RAM 64 or more etc, 10Mb Ethernet adapter.

- To drive the CAVE application: any Unix-based system capable of running CAVE applications.

Software Requirements

- IBM Via Voice Pro Speech Recognition Software

- jdk 1.2 or higher version

- CAVE Virtual Reality library

- CAVERNsoft G2 Tele-Immersion library

 

Instructions for running the demo application

Follow these steps to setup CAVERNsoft Voice Recognition

This setup assume that CAVERNsoft is already install

Via Voice

- Install Via Voice on PC if not installed.

- Setup the microphone as Via Voice instructed.

- For best performace each user must create a personal voide Model.

CAVERNsoft Java Interface

- Download CAVERNsoft Java Interface to PC

- Install the CAVERNsoft Java Interface

- Make a directory CJavaIntface

- Unjar the CJavaInterface.jar file in CJavaIntface directory.

- Type "jar vxf CJavaInterface.jar " on the command line

- It will make lib, include, bin directory

- Change the gramma

-To run it go to the bin directory and type "cavern" on the command line

CAVERN_voiceRecog_c usage

- Include CAVERN_voiceRecog_c in to you c++ program i.e. Sample

- Compile and link the program using Makefile

- Sample program use CAVE Simulator, in this case, on SGI the left button of the mouse will put the CAVERN_voiceRecog_c object to sleep by setting status field to 0 and will send a signal to Java Server on the PC to sleep. Right button will set the status field to 1 and will send a signal to Java Server on the PC, to start.

Note: Make sure to change the server name in you program or in Simple program (voiceSample.c++).

Customizing for your own applications

- Changing the grammar File