Data processing: speech signal processing – linguistics – language – Speech signal processing – Recognition
Reexamination Certificate
1999-03-22
2001-08-07
Dorvil, Richemond (Department: 2741)
Data processing: speech signal processing, linguistics, language
Speech signal processing
Recognition
C704S246000, C704S278000
Reexamination Certificate
active
06272461
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to a computer-based presentation support system, and more particularly to a system that provides visual aid to a presenter during the course of a presentation.
DESCRIPTION OF THE RELATED ART
Verbally communicating prepared information to a live audience can be quite challenging. As an alternative to memorizing long speeches and/or detailed information, visual aids are utilized by presenters to guide them through a presentation. Visual aids may include a printed copy of a speech that is held by the presenter, cue cards flipped by a support person, or a teleprompter that is manually paced by a support person. With advances in computer technology, especially desktop and laptop computers, visual aids may also include an electronic copy of a speech that is displayed on a computer screen to a presenter such that the speech can be directly read from the screen by the presenter. In all cases, the presenter or the support person must manually progress through the visual aid by either turning the pages of a printed speech or scrolling through screens of an electronic copy of a speech. Requiring the presenter to manually turn pages or scroll through computer screens may take the presenter's focus away from the content of the presentation or may distract the presenter from connecting with the audience. Moreover, requiring a support person to move through prepared information by flipping cue cards or controlling a teleprompter is costly.
There are known techniques for synchronizing recorded verbal information with prepared textual information, however, the techniques do not provide benefits to a live presenter. An example technique is disclosed in U.S. Pat. No. 5,649,060 entitled “Automatic Indexing and Aligning of Audio and Text Using Speech Recognition,” issued to Ellozy et al. (hereinafter Ellozy). Ellozy discloses a computer-based system that synchronizes and indexes recorded verbal information with an electronic transcript of the recorded verbal information. An example application of Ellozy is synchronizing and indexing a transcript of video recorded testimony with the video recording. The system of Ellozy inputs the recorded verbal information into a speech recognizer in order to transform the recorded verbal information into decoded text. The decoded text is then automatically matched with the original electronic transcript via identification of similar words or clusters of words. By matching the decoded text with the original written transcript, an entire written transcript can be indexed with the recorded verbal information, so that a portion of recorded verbal information can be easily located through a text search of the electronic transcript. While Ellozy provides synchronization between recorded spoken words and written text, Ellozy does not disclose a system that is responsive to spoken words in real-time. In addition, while the Ellozy system may be used as a conventional teleprompter, allowing a presenter to read a written transcript from a display screen, the Ellozy system would require manual control of the pace with which the written transcript is scrolled through the teleprompter, thereby providing no benefits over the above-mentioned presentation support techniques.
As a result of the shortcomings of known techniques for providing support to presenters, what is needed is a computer-based technique that has the intelligence to automatically provide visual support to a presenter in a timely manner without physical interaction by the presenter or any other support person.
SUMMARY OF THE INVENTION
A method and an apparatus for providing visual aid to a presenter involve converting the spoken words of a presenter into an electronic text format, electronically comparing the converted spoken words to electronically stored reference text to find text string matches, utilizing the text string matches between the converted spoken words and the reference text to determine a current location of the presentation with respect to the reference text, and delivering upcoming portions of the reference text to the presenter as needed to enable a continuous presentation by the presenter. In a preferred embodiment a presentation support system incorporated into a personal computer includes reference text storage, a speech capture subsystem, a speech recognition subsystem, a comparator subsystem, a processor, a speech pace monitor, and a display. Some of the units in the presentation support system include dedicated equipment, while other units in the presentation support system are associated with standard personal computer equipment.
The reference text storage includes rapidly accessible computer memory for storing reference text. The reference text stored in the memory reflects the prepared information that is to be presented by the presenter. The reference text may be in the form of, for example, a verbatim copy of a speech, an outline of presentation content, or a bullet list of talking points. The reference text may be input into the computer memory through various techniques, including typing the text into a computer-based word processing program.
The speech capture device captures the sounds generated when the presenter speaks and converts the sounds into electronic signals that are representative of the words that are spoken by the presenter. The speech capture device is preferably a microphone system that is incorporated into the personal computer.
The speech recognition system converts the electronic signals from the speech capture device into an electronic text format. The speech recognition system is powered by hardware and software that enable spoken words to be converted into electronic text in real-time, or with negligible delay after the words are spoken by the presenter.
The comparator correlates the converted electronic text to the reference text in order to identify matching text strings between the spoken words and the reference text. The matching text strings may involve a single word, but preferably involve multiple words in order to maximize the likelihood that the spoken words correspond to the appropriate section within the reference text.
The speech pace monitor determines the pace at which the presenter is speaking. The pace of speaking is preferably measured by calculating the number of words per unit of time that are being spoken by the presenter. The pace can be determined by monitoring actual words spoken or monitoring the progression of a presentation through the reference text.
The processor determines the content and timing of information that is to be provided to the presenter. The processor preferably utilizes realtime information concerning the location of the spoken words relative to the reference text to determine which portion of the reference text will be needed by the presenter next. In a most preferred embodiment, the processor also utilizes speech pace information from the speech pace monitor in order to appropriately time the delivery of reference text to the presenter. Determining the content and timing of information that is provided to the presenter is performed by the processor on a continuous basis, such that adjustments to the delivery of reference text can be made throughout a presentation.
The monitor is preferably a conventional computer monitor or screen that enables the reference text to be read by the presenter. The monitor is preferably positioned so that it is viewable by the presenter and not by the audience. By discretely providing reference text only to the presenter, the audience is not distracted from the message of the presenter.
In operation, reference text is prepared and stored in the reference text memory of a personal computer before a presentation is scheduled to begin. With the reference text stored in the personal computer memory, the presentation support system is activated prior to delivery of the desired presentation by the presenter.
Once the presenter begins speaking, the presenter's spoken words are captured by the speech capture subsystem and t
Aktas Christoph A.
Meredith Phillip C.
Abebe Daniel
Dorvil Richemond
Siemens Information and Communication Networks Inc.
LandOfFree
Method and apparatus for an enhanced presentation aid does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for an enhanced presentation aid, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for an enhanced presentation aid will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2442938