edu.ucsb.nmsl.autocap
Class AutoCaptioner

java.lang.Object
  extended by edu.ucsb.nmsl.autocap.AutoCaptioner

public class AutoCaptioner
extends java.lang.Object

This class contains all the logic for AutoCap. AutoCap is an application for the automatic alignment of a segmented transcript with a video for the purpose of providing captions. This class implements the needed event handlers for the Sphinx Speech recognition system for the first process AutoCaptioner the aligns the recognized with the text from the input transcript. After this alignment phase is complete it is necessary to estimate the time-stamps for all captions for which the first word was not recognized. Estimation is accomplished during the estimation phase.

Version:
1.0
Author:
Allan Knight
See Also:
"AutoCap: Automatic Captioning of Mutlimedia Presentations"

Field Summary
(package private)  CaptionAligner aligner
          The CaptionAligner instance that will align all the captions with the in put video.
(package private)  boolean transcribeOnly
          Indicates whether alignment is done, or just transcription.
 
Constructor Summary
AutoCaptioner()
          This default constructor creates a default instance of AutoCaptioner.
 
Method Summary
static void main(java.lang.String[] args)
          This method is the main method for running the AutoCap application.
 void reportResearchStatistics(Transcript t)
          This method reports the statistics used for research.
 void reportStatistics(Transcript t)
          This method reports the statistics used for the user interface.
 void setTranscribeOnly(boolean t)
          This method is used to set parameter transcribeOnly.
 void start(java.lang.String media, java.lang.String xml)
          This method is called from main to start recognizing a file.
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Field Detail

aligner

CaptionAligner aligner
The CaptionAligner instance that will align all the captions with the in put video.


transcribeOnly

boolean transcribeOnly
Indicates whether alignment is done, or just transcription.

Constructor Detail

AutoCaptioner

public AutoCaptioner()
This default constructor creates a default instance of AutoCaptioner.

Method Detail

setTranscribeOnly

public void setTranscribeOnly(boolean t)
This method is used to set parameter transcribeOnly. If true, the program only transcribes a video rather than align a transcript.

Parameters:
t - - Boolean indicating whether to turn of transcriptionOnly or not.

start

public void start(java.lang.String media,
                  java.lang.String xml)
This method is called from main to start recognizing a file.

Parameters:
media - The file location of the media that is to be aligned with its transcript.
xml - The XML file containing the transcript. See the documentation for the class QADTranscriptFileWriter for details on the file format for captions used in AutoCap.

reportResearchStatistics

public void reportResearchStatistics(Transcript t)
This method reports the statistics used for research. This method is not necessary for normal opereation of AutoCap.

Parameters:
t - The input transcript of the audio portion of the input video.

reportStatistics

public void reportStatistics(Transcript t)
This method reports the statistics used for the user interface. This method is not critical for the normal operation of AutoCap, but does help the user understand how well AutoCap performed during a particular running.

Parameters:
t - The input transcript of the audio portion of the input video.

main

public static void main(java.lang.String[] args)
This method is the main method for running the AutoCap application. It is the only necessary main method in any of the classes that make up the AutoCap application.

Parameters:
args - Array of strings passed from the command line. args[0] contains the name of the media file that contains human speeh. args[1] the name of the XML file that contains the transcript.