US 5,006,987 AGrant
Audiovisual System for Simulation of an Interaction Between Persons Through Output of Stored Dramatic Scenes in Response to User Vocal Input
Issue Date:1991-04-09
•29 Claims
•2 Drawing Sheets
Abstract
A method and apparatus for producing an interactive video drama employing a personal computer, a voice recognition device, and a programmable videodisc player. A plurality of dramatic scenes are stored on a videodisc and displayed to the user in response to spoken commands. At certain scenes designated as decision points, the next displayed scene is selected based on a probability technique. By storing audiovisual representations of a speaking human being and by selecting scenes for display based on spoken input signals, the apparatus provides the illusion of dialog between a user and prerecorded human characters.
Metadata
Inventor
- William G. Harless
Application Information
Application Number:US 8435529
Filing Date:1986-03-25
Priority Date:1986-03-25
Art Unit:238
Classifications
IPC:
G06F 1538
Field of Search:
364381434358513.5;41943307-309;321;335342
Patent Drawings (2 sheets)
Description
Background of the Invention
The invention relates to an interactive audiovisual system and, more particularly, to an interactive audiovisual system employing a voice recognition unit and a programmable videodisc player.
There is an increasing need in various fields of education, and particularly in professional education, to provide learning opportunities in a realistic environment without the risk of harm caused by improper student action. For example, it is extremely desirable for medical students to interact with patients at an early stage in their educational process to develop practical experience and provide a foundation to which the students' academic education can be related. Although traditional forms of medical education allow students to interact with actual patients under the supervision of a physician, in order to avoid risk to the patient the medical student is prohibited from directly making independent decisions involving treatment of the patient.
Various forms of simulation of the student-patient interaction are known in the prior art. For example, existing systems are known which employ a programmable videodisc (player) under the control of a personal computer to receive student input commands through a keyboard or touch-screen CRT input device. Although existing systems can often provide useful educational experiences for the student, it is desirable to provide a more natural means of interaction for the student so as to increase the realism and maximize the effectiveness of the educational method. It is also desirable to provide a continuous audiovisual display of stored dramatic scenes to heighten the illusion of an actual physician-patient relationship. It is further desirable to increase the realism of an interactive audiovisual system such that the outcome of the displayed audiovisual drama will be influenced by, but not completely determined by, student input to the system.
Summary of the Invention
The invention achieves the aforementioned objectives by providing, in one aspect, an interactive audiovisual system comprising means for storing a plurality of prerecorded audio and video scene signals representative of a speaking human being; means for receiving voice signals and for generating interrupt signals representative of the meaning of words in said voice signals; and means for selectively displaying specific ones of said scene signals in response to said interrupt signals to provide an audiovisual representation of a human being speaking an appropriate response to said received voice signals.
In another aspect, the invention provides a method for providing an interactive drama, comprising the steps of storing a plurality of prerecorded audio and video scene signals representative of a speaking human being: receiving voice signals and generating interrupt signals representative of the meaning of words in said voice signals; and selectively displaying specific ones of said scene signals in response to said interrupt signals to provide an audiovisual representation of a human being speaking an appropriate response to said received voice signals.
Brief Description of the Drawings
FIG. 1 is a block diagram of apparatus constituting a preferred embodiment of the invention; and
FIG. 2 is a flow diagram of an interactive drama illustrating the principles of the preferred embodiment of the invention.
Description of the Preferred Embodiments
Reference will now be made in detail to the present preferred embodiment of the invention, an example of which is illustrated in the accompanying drawings and in which like reference characters refer to corresponding elements.
FIG. 1 is a block diagram of a preferred embodiment of the present invention. As shown in FIG. 1, a computer system 10 includes a computer 12 having a mass memory device 14 and a keyboard 16. In the preferred embodiment, computer 12, mass memory 14, and keyboard 16 are provided by an IBM PC/XT personal computer. A videodisc player 18 is connected through a serial port and RS-232-C connector to computer 12. In the preferred embodiment, videodisc player 18 may be a Pioneer model LD-V6000 programmable videodisc player.
A PC bus expansion card 20 is connected to computer 12. Also connected to computer 12 is a voice recognition and digitized speech unit 22 which includes a microphone 24. Voice recognition and digitized speech unit 22 may be a commercially available product such as the Votan VPC 2000 speech processor card.
A graphics adaptor 26 is also connected to computer 12 and may be, for example, the VAL and Tecmar graphics adaptor combination The output of graphics adapter 26 is supplied to a monitor 28.
The specific components described above can be incorporated into system 10 and computer 12 can be programmed using well-known methods to provide the features and functions to be described hereinafter. However, in the preferred embodiment the functions of PC Bus and expansion card 20, voice recognition and digitized speech unit 22 and graphics adaptor 26 are provided in an integrated computer expansion chassis 30 which is part of an IMSATT model 2000 computer controlled videodisc system. This equipment is commercially available from the IMSATT Corporation of Falls Church, Virginia.
Videodisc player 18 is a read-only device which supplies audio and video signals to monitor 28 from a preprogrammed readonly removable disc. The disc may contain up to 54,000 video still frames or approximately 30 minutes of continuous video information. The disc also includes two tracks of audio associated with the recorded video data. Video disc player 18 may provide single frame or continuous video images to monitor 28 as commanded by computer 12 which transmits address and control commands through a serial port and RS232-C connection to videodisc player 18 to select the desired audio and video signals.
Voice recognition and digitized speech unit 22 provides several capabilities. First of all, voice signals supplied through microphone 24 are interpreted by unit 22 in conjunction with computer 12 to provide specific interrupt signals related to desired words. For example, unit 22 may be programmed to recognize up to ten sets of 75 separate voice patterns. That is, the voice signal representative of the word "begin" may be "taught" to unit 22 such that when a user speaks the word "begin" into microphone 24, unit 22 will generate a unique interrupt signal to computer 12. Each set of 75 words is known as a "vocabulary" and may be stored in computer memory 12 to be switched instantaneously.
Unit 22 also provides the capability to generate human speech by storing digital signals through computer 12 into mass memory 14 in a compressed form such that, under program control, computer 12 may retrieve the digital speech signal from unit 14 and command unit 22 to generate signals to monitor 28 which represent preprogrammed desired voice signals.
Specific connections of videodisc player 18, and unit 30 representing the functions of PC Bus and expansion card, voice recognition and digitized speech unit 22, and graphics adaptor 26 are provided in the IMSATT 2000 User Manual, Version 1.8, February 1986, commercially available from the IMSATT Corporation. Additional details of the voice recognition and digitized speech capability of unit 22 are provided in the Votan VPC 2000 Users Guide, commercially available from the Votan Corporation of Freemont, California. The disclosures contained in these manuals are hereby expressly incorporated by reference.
The invention thus includes means for storing a plurality of prerecorded audio and video scene signals representative of a speaking human being. In the preferred embodiment, such means comprises programmable videodisc player 18.
The invention also includes means for receiving voice signals and for generating interrupt signals representative of the meaning of words in the voice signals. As embodied herein, such means include microphone 24 and voice recognition and digitized speech unit 22.
The invention further includes means for selectively displaying specific ones of the scene signals in response to interrupt signals to provide an audiovisual representation of a human being speaking an appropriate response to the received voice signals. In the preferred embodiment, such means includes computer 12 and monitor 28.
Referring now to FIG. 2, there is shown a flow diagram of an interactive drama produced using the apparatus and methods of a preferred embodiment of the present invention.
The material required to implement such a video drama consists of a videodisc containing the scenes (motion and still frame) which comprise the story and the characters involved in the drama. This videodisc is the product of a process involving:
1. program planning and design,
2. scripting and content evaluation,
3. auditioning and actor selection,
4. staging and videotaping,
5. tape editing and videodisc pressing.
The resulting videodisc contains color video signals comprising some combination of the approximately 30 minutes of motion or 54,000 still frames, and two channels (approximately 30 minutes each) of audio. Each frame on the disc is addressable and the maximum search time is 3 seconds.
A computer program and related files and graphics are generated to play the disc and establish the interactivity necessary for the drama. The IMSATT development system (commercially available) is the preferred embodiment for this task. This system is described in the previously referenced IMSATT User Manual.
In order to generate the program, disc catalogs and design documentation (additional information useful for programming) must be prepared. The catalog provides the raw data for the computer program; It describes the location of each video scene on the disc by in-and-out-frame number and the spoken voice commands associated with each scene. It also describes graphic overlay and digitized speech requirements when appropriate to a scene. Appendix 1 contains the disc catalogs and design documentation required for generating the interactive drama which will be described below. Appendix 2 contains specific examples of the program which was developed using the IMSATT system and the previously mentioned catalogs and design documentation.
The interactive medical drama of FIG. 2 portrays a situation in the life of a fictitious patient named Frank Hall, who is played by a professional actor. Frank Hall is a 46-year-old man who comes to the emergency room complaining of weakness and abdominal pain following an episode of vomiting blood two days earlier. The simulated patient illustrates the medical content and sociological issues related to a variety of gastroenterology problems confounded by alcoholism.
The Frank Hall drama contains a variety of scenes recorded on a videodisc as scene signals depicting the patient's current experience in the hospital and significant past experiences (through audiovisual flashbacks) related to his present illness. Other scenes depict elements of the physical examination and relevant diagnostic tests and procedures with visuals of x-rays, endoscopic views and so forth, whenever appropriate.
The user (the instructor, in concert with students) assumes the role of the physician in this interactive drama and, through spoken inquiries and commands, interviews the simulated patient and directs the course of action. The system responds instantly to each spoken inquiry and/or command.
Before initiating the drama, the instructor trains the system to recognize 140 spoken command words consisting of medical history concepts (presenting complaint, pain, smoking habits, drinking habits, etc.) and terminology (hematocrit, urinalysis, endoscopy, etc.) to diagnose and manage a patient. A list of these spoken command words is contained in the design documentation of Appendix 1. The process for training takes approximately thirty minutes and is described in detail in the aforementioned IMSATT User Manual. The computer records the way the instructor says each spoken command word in the form of a digitized voice pattern. These patterns are stored and, subsequently, used by the computer to recognize the instructor's inquiries and commands during the presentation, as described in the aforementioned Votan Users Guide. The current Votan system restricts the number of words available at one time to 75, plus synonyms. However, memory switching of a set of vocabulary words during the drama can be accomplished in less than one second, providing virtually unlimited vocabulary.
Once the instructor has trained the system, he/she can act as the teacher, guide, facilitator and/or spokesperson for the class.
Following the start up of the computer in the conventional manner, a still frame of the character, Frank Hall, is presented on monitor 28 with a graphic over title, "The Case of Frank Hall." The system waits for the user to begin the session.
The user speaks the command "begin" into microphone 24. The drama begins with a full color action scene, including background music, of Frank Hall driving a large, battered automobile into the hospital parking lot. The scene changes to a disheveled Frank Hall at the emergency room reception desk.
Frank Hall (to the emergency room (ER) nurse): "I'm sick; I need to see a doctor."
A dialog between Frank and the nurse takes place for approximately 30 seconds; then the action on monitor 28 freezes. This freeze in the action is called a "wait state." It serves as a subtle prompt to the user to take charge and do something.
There are no menus or cues in the interactive drama. The user must rely on personal resources and initiative to determine solutions and courses of action. Meanwhile, the system loads the vocabulary patterns for the dialog section (block 200, FIG. 2) and remains in wait state until a verbal command is given.
To see the patient, the instructor may say, "Mr. Hall," which brings the patient, full face to the screen where he sits looking directly at the class, waiting to respond to questions. Alternatively, the instructor may say, "start the workup," which elicits another action scene of Frank Hall in an examining room having his vital signs taken by the emergency room (ER) nurse. This scene changes to one with the patient in an examining gown, again looking directly at the group and waiting to answer questions. These, too, are wait states where the instructor and students must determine the course of action. In this example they decide to interview the patient.
User: "What's your problem?"
Frank Hall: (looking directly out at the student) "I'm sick. Ive been throwing up blood . . . I'm afraid I might have cancer . . . " (Wait state.)
User: "When did it begin?"
Frank Hall: "Well, it started last week. I was in this restaurant . . . "
The scene becomes a flashback to the restaurant showing Frank Hall talking to the waitress and beginning to feel sick. The patient continues to describe the situation as the flashback scene changes to his bathroom and an episode of vomiting blood; it concludes back in the examining room in a wait state. This user-controlled dialog with the simulated patient can continue through a complete social and medical history or the user may digress to request test procedures or other information.
User "Vital signs."
Values for Frank Hall's temperature, blood pressure, pulse and respiration appear as text over a scene of the nurse taking his vital signs.
To continue talking with the patient, the user simply says the patient's name and he reappears on the screen.
User: "Mr. Hall."
(Frank Hall reappears).
Frank Hall: "Yes."
User: "You mentioned your wife."
Frank Hall: "Rita's my wife . . . er, she was my wife. We're separated now . . . She's filing for divorce. We've been married for 15 years."
The interactive drama allows the class to probe deeper into history concepts to elicit more information from the patient. The user may do this by asking open-ended questions such as: "Is there more?", "Tell me more", "Is that all?" Probing questions concerning Frank Hall's marriage reveal, through flashback, a wife beating incident and other pertinent information about the patient and his relationships.
The patient will continue to respond until he has exhausted his responses about that concept at which time he will say that there is nothing more to say. Further probing by the user about that subject will cause the system to randomly select a generic scene of Frank Hall indicating that no further information is available. This generic response may indicate passiveness, impatience, or other human emotions. This capability provides a realistic uncertainty concerning the character's emotional response to probing that creates dramatic tension and piques interest. The user may also change the subject and later return to it to get additional information. For example, the user may ask about concept A (e.g., smoking habits) and get a response (#A1); then ask about concept B (e.g., drinking habits) and get a response (#B1); then ask for more information about smoking (concept A). The drama, having recorded the display of scene #A1, will not repeat scene #A1 but will select #A2 (or An), the next of the dramatic scenes associated with the spoken command "smoking". All such scenes associated with a specific concept are contiguously recorded on the videodisc so that they can be displayed in a seamless manner, that is, without a blackout which would otherwise occur as a result of the videodisc player initiating a search.
Thus, probing questions never elicit a redundant response from the patient. This feature enhances the believability of the drama and precludes rigid contrivance which has characterized computerized simulations in the past.
This interactivity using natural language (spoken English) voice commands occurs in the dialog section of the drama (block 200). This important section allows the user the capability to engage in direct dialog with the "talking head" of the simulated patient; to view flashbacks at specific times during that dialog, which reveal incidents and aspects of the patient's life that are relevant to the problem at hand; to probe about any concept in the simulated patient's memory without viewing the same scene twice; and, to view immediate textual responses to requests for information about physical examination and diagnostic tests. The interactive drama provides the very powerful illusion of natural conversation between internal character(s) in the video drama and the external character (i.e., the user). The user may act as the problem solver in the drama and, in that role, converse with the internal character(s) and spontaneously intervene in the drama to influence the outcome(s) of the situation being portrayed.
The dialog with the patient will continue until the user either: (1) admits the patient to the hospital (block 202); or (2) dismisses the patient from the hospital (block 203). Should the user speak the command "admit the patient" (block 202), a decision point in the drama is reached.
A decision point is a scene which can lead to any of a plurality of possible outcome scenes. The specific outcome scene which is displayed following a decision point scene is determined, in the present invention, by a probability method. Each outcome scene has associated with it a predetermined probability factor. A random number is generated and compared to the probability factors for each outcome scene. The outcome scene whose probability factor is matched by the random number is the scene which will next be presented to monitor 28.
The probability factors are not unchangeable. Rather, they can be altered by influence factors, that is, specific voice commands entered by the user (all of which are recorded by the system). For example, if the user speaks the command "abdominal exam" this increases the probability that decision point 1 will be followed by outcome 1.1 Admission.
The system generates a random number which is compared to the probability ranges associated with decision point 1 (FIG. 2). Outcomes of decision points are probabilistic rather than deterministic and, therefore, more accurately reflect the real situation. At each decision point in the model there is a probability-based decision table which contains the consensus opinion of experts concerning the level of likelihood of each possible outcome for that point of decision. An outcome may be determined to be unlikely, likely, or very likely depending upon the combination of events which have occurred at the point of decision. For example, if the user has performed a complete abdominal examination for Frank Hall in the examination room, the likelihood of admission will increase. If he has taken a complete medical history as well as an adequate physical examination the likelihood of admission will further increase, etc. Each of the levels of likelihood (very likely, likely and unlikely) has an associated numerical range which is used in comparison with a computer generated random number to determine the outcome in question. The precise numerical range for any likelihood level is dynamically determined during the course of the drama, in a manner to be more completely described. A detailed description of the probability method used in the interactive drama is attached as Appendix 3.
If the user speaks the command "admit the patient," and if the outcome of decision point 1 is Admission (block 204), then vocabularies are switched at block 204 and a dialog ensues between the user and a prerecorded voice from monitor 28. The user is interrogated by the monitor concerning perceived differential diagnoses and desired diagnostic tests for the patient. The system confirms each of the user's verbal entries with onscreen text in the following manner.
User: "Admit the patient."
Monitor voice: "Please state your differential diagnoses, one at a time."
User: "Hemorrhagic gastritis."
Monitor voice: "Is there another?"
User: "Mallory-Weiss Syndrome."
Monitor voice: "Another?"
User: "Cancer."
Monitor voice: "Is there more?"
User: "That's all."
The monitor voice then asks the user to provide, in the same manner, a plan for diagnostic workup and for immediate management.
The system schedules the diagnostic procedures ordered by the user. The system will schedule only one invasive test (liver biopsy, endoscopy, etc.) per day, and no more than three noninvasive tests per day. The results for a test will not be provided until an appropriate amount of time has passed since that test was ordered.
The hospital stay (HS) scenes (Appendix 1) will begin to play sequentially. The instructor and students may simply observe these scenes, watching the patient as he goes through each day, or stop the action by appropriate voice commands to revise the diagnostic workup and/or therapeutic plan. The new requests will be noted and the results of these tests will be available in graphics text form upon command after an appropriate amount of time has passed. The flow of hospital stay scenes will resume when the user commands "continue".
At any time during any scene, the user may use the command word "thoughts" to hear what the patient is thinking. The purpose of this feature is to remind students to be aware of the stresses that patients are coping with, which more often than not are never verbalized and thus remain unknown to the physician. This raises the students' consciousness concerning the patient's mental/emotional state and broaden their perspective of the patient's clinical experience.
"Thoughts" are considered to be remembrances, ideas, and/or concerns of the patient which are not elicited by direct inquiry during the course of the interaction. These thoughts may be auditory (inner voices) or visual (images). Thoughts that are auditory are recorded and stored as scene signals in the form of digital audio in an addressable file on the mass storage device 14 and, when retrieved, played from monitor 28 as human speech signals. Visual thoughts are stored as addressable images, i.e. scene signals on the videodisc. Thoughts that are relative to a playing scene are flagged as such and will be presented if the user speaks the command "thoughts" when that scene is playing. If thoughts are commanded and there are no "thoughts" specifically related to the scene being played, a visual thought is randomly selected from the videodisc or an auditory thought from the mass storage 14 and those scene signals are presented to monitor 28.
Also, the system keeps track of those flashbacks associated with scenes in the dialog section which were not previously presented and assigns such scenes to the random thoughts video database.
Following two days of simulated time in the hospital, decision point 4 (block 207) is reached, and a decision is made by the system through the probability method as to whether there will be a crisis. That is, a random number is generated and compared to the outcome possibilities of decision point 4 of FIG. 2. Outcome 4.1 allows for no crisis to occur and the hospital stay continues (block 208) until the user decides to discharge the patient.
A life threatening crisis (outcome 4.3) occurs if the random number generated for that decision point is in the range associated with the crisis outcome, and a scene is displayed in which Frank Hall vomits blood in the early morning hours while the user is looking on. Tension is created by the hysteria of the roommate. Meanwhile the system loads a vocabulary which allows the user to intervene (or choose not to intervene). This is block 222 of FIG. 2. Should the user not intervene, decision point 6 is reached and Frank Hall either expires (outcome 6.1) or survives (outcome 6.2). If he expires, then the drama ends and feedback is available to the user. If he survives, the aftermath of the crisis occurs with scenes displayed of the nurse doing generic nursing while the user speaks commands to request management procedures and diagnostic tests which are appropriate for the situation.
Decision point 7 comes into play by the passage of time to determine if Frank Hall will stabilize or not. Outcome 7.1 shows the nurse taking over due to perceived inadequacies of the user/physician. Outcome 7.2 shows Frank Hall going into shock following his loss of blood. Outcome 7.3 shows Frank Hall stabilizing. All of these outcomes of decision point 7 lead back to the hospital stay section, which is shown as 208 on FIG. 2. Should the user intervene at block 222 during the crisis, decision point 5 comes into play to determine if Frank Hall survives or expires. Outcome 5.2 shows Frank Hall expiring, leading to the feedback for the user and the end of the drama. Outcome 5.1 shows Frank Hall surviving and leading into the previously described decision point 7.
The capability of the system to make dynamic unpredictable decisions about certain situations provides a life-like mystery to the drama and offers the opportunity for discovery learning to occur in the classroom. Even the instructor does not know the direction the drama will take nor the outcome of any event.
Ultimately, if Frank Hall survives the crisis, the command word "discharge the patient" will stop the process at block 210 and the system will switch vocabularies, display a freeze frame, and interrogate the user concerning final diagnoses and discharge plan.
Having obtained the discharge plan by spoken commands from the user, the system determines the outcome for the patient by generating a random number and comparing it to the probability values associated with decision point 8. Outcome 8.1 will display a scene where Frank Hall dies. Outcome 8.2 shows a recovery and ultimate positive outcome for Frank Hall. Outcome 8.3 shows Frank Hall as a revolving door patient; that is, a patient who returns to the hospital with essentially the same health problems over and over again.
Each of these outcomes is followed by the feedback section. When the command "feedback" is spoken by the user, performance feedback is provided. The user is scored based on the amount of critical information obtained from Frank Hall, the appropriateness of diagnostic tests and management ordered and hospital cost and medical cost imposed on the patient. These and other feedback parameters are available from a record which is maintained by the system of each scene which has been played during the drama. Feedback is block 212 in FIG. 2.
Referring back to decision point 1 of FIG. 2, outcome 1.2 is a referral where a determination is made that the patient cannot be admitted due to lack of money and/or insurance. The user is apprised of this through a monitor voice at block 220 in FIG. 2.
Referral also provides a freeze frame wait state for the user to decide whether or not he or she will accept referral as an alternative to admission, or whether the user will once again speak the command "admit the patient" to request admission. Acceptance of the referral alternative by the user initiates a decision at decision point 3 to determine the outcome for the patient following referral. Outcome 3.1 provides the user with information that the patient was lost to follow-up. Outcome 3.2 allows the patient to be admitted to County Hospital, which accommodates indigent patients. Outcome 3.3 provides information back to the user that Frank Hall expired. Each of these outcomes is followed by the feedback section, and the drama ends.
Referring back to decision point 1, outcome 1.3 provides a monitor voice informing the user that admission is not allowed and suggesting Frank Hall be treated as an outpatient. Acceptance of this alternative causes the system to switch vocabularies in order to interrogate the user concerning his or her diagnosis and management plan for the patient on an outpatient basis. This is block 232 in FIG. 2.
Following the submission of the final diagnosis and management plan by the user, the system determines the outcome for Frank Hall utilizing a probability method at decision point 2. Outcome 2.1 informs the user that Frank Hall expired. Outcome 2.2 informs the user that Frank Hall recovered and became gainfully employed. Outcome 2.3 shows Frank Hall as a revolving door patient in the hospital. Outcome 2.4 informs the user that Frank Hall went off, got sick, and decided to litigate for malpractice. Each of these outcomes takes the user to the feedback section (block 212).
The interactive drama ends following the outcome scene. At this point the instructor may request feedback from the system, which is presented in the form of charts and graphs describing the group's performance with the patient. The feedback parameters include:
1. correctness of final diagnoses;
2. correctness of discharge plan;
3. percentage of critical information obtained during interaction with the case (according to expert committee), with breakdown;
4. danger/discomfort index for tests and procedures ordered, with breakdown;
5. costs incurred by the patient based on procedures ordered and length of hospital stay;
6. appropriateness and effectiveness of crisis intervention.
These feedback features are implemented by the continuous storing of scene signals which were displayed to the user in response to the user's uncued commands throughout the drama.
Specific features of the invention will now be described in greater detail.
There are eight internal decision points in the drama just described each having multiple associated outcomes. For every decision point there is an associated probability table which contains the influence factors that determine the liklihood levels (likely, unlikely, very likely) for each of the outcomes. These probability tables are included in Appendix 2.
Each decision point scene has multiple associated outcome scenes, one of which is selected by the computer by means of a probability procedure. Each decision point scene is identified and stored by its specific address on the videodisc. The associated outcome scenes are also identified and stored by disc address.
When a decision point is reached during the course of a user interaction, the probability procedure is initiated in the following manner:
The record of scenes displayed and spoken commands is consulted to determine which influence factors have been entered. As an example, reference may be made to decision point 7 of FIG. 2. That is, the patient Frank Hall has survived a crisis, but still faces uncertainty over his ultimate fate. A scene is displayed of the patient lying in bed with a nurse attempting to stabilize his condition. The possible outcome scenes from decision point 7 are:
7.1 The nurse takes over and orders the proper procedure (i.e., start appropriate medication)
7.2 Frank Hall goes into shock and dies.
7.3 Frank Hall stabilizes.
The probabilities of these possible outcomes are influenced by the type of medical procedures which the user has previously ordered by spoken commands. This is shown in Decision Point Table 7 of Appendix 2 which lists the likelihood levels (VL-very likely, UL-unlikely, L-likely) of each outcome for each combination influence factors. The influence factors for decision point 7 are: intravenous (IV) medication ordered, vital signs checked, and transfusion ordered.
It can be seen at influence factor combination address 4 that if no IV has been ordered, if vital signs have not been checked, and if a transfusion has been ordered, it is likely that the nurse will take over, unlikely that the patient will die, and very likely that he will survive. These estimates are entered into the system by experts during development and may be changed by the instructor prior to using the system.
When decision point 7 is reached during the interactive drama, the system converts the L, UL, VL operator-entered likelihood levels to probability ranges V.sub.1, V.sub.2, V.sub.3 equal to 33-66, 0-32, 67-99 respectively. Next, a random number is generated within the range of each V value, for example V.sub.1 =48, V.sub.2 =5, and V.sub.3 =93. Next, the system computes probability ratios for each V value. That is, the probabilities are normalized as shown below and described in Appendix 3. ##EQU1## P' ranges for each outcome are then established: ##EQU2##
A random number between 0 and 99 is then generated and compared to the P'probability ranges to determine which associated outcome scene O.sub.c is chosen. The disc address for the selected outcome scene P'.sub.1, P'.sub.2 and P'.sub.n is determined from scene attribute information (generated from disc catalogs and stored in the system) and the selected outcome scene is displayed on the monitor.
Of the dramatic scenes which are stored on the videodisc for a selective display, certain types are of particular interest, as noted below:
Type 1--A scene for which a specific relevant "thoughts" scene is associated. For example, a scene depicting an uncomfortable diagnostic procedure is a type one scene which has associated with it a "thoughts" scene (audio only) expressing pain.
Type 2--A scene for which there is no specifically associated "thoughts" scene. An example of a type two scene is Frank Hall strolling through the hospital gift shop.
Type 3--A "thoughts" scene which is directly associated with a type one scene. An example of a type three scene is the scene expressing pain, noted above.
Type 4--A "thoughts" scene which is not directly related to any type one scene, but which may be displayed in association with a type two scene. Examples of type four scenes are scenes expressing boredom, flashbacks to a previous portion of Frank Hall's life, and a scene expressing concern with the total cost of the hospital stay.
Type 5--A scene of a dialog relating to a specific question or concept. Examples of type five scenes are those in which Frank Hall relates information concerning his marriage. There may be several type five scenes associated with each concept. For example, repeated questions to the character inquiring about his marriage will result in sequential display of a plurality of type five scenes relating different aspects of the character's marriage. All type 5 scenes associated with a specific concept are arranged contiguously on the videodisc to eliminate the need for a videodisc player 18 search which would cause a brief blackout of monitor 28.
Type 6--A generic dialog scene stating that there is no additional information which the character can supply in response to the particular concept. For example, when all of the type five scenes containing information relating to the character's marriage have been displayed, an additional inquiry into marriage will result in the display of a type six scene in which the character states, "I can't tell you any more about that" or "That's all there is."
The invention thus includes means for designating scene signals as type one, type two, type three or type four scene signals and for associating each of the type three scene signals with one of the type one scene signals and for associating each of the type four scene signals with all of the type two scene signals. In the preferred embodiment, such means includes computer 12 programmed in the manner specified.
The invention further includes logic means for selectively presenting specific ones of the scene signals to display means in response to interrupt voice signals, for responding to the interrupt signals during presentation of the type one scene signals to initiate presentation of one of the associated type three scene signals and for responding to the interrupt signals during presentation of the type two scene signals to initiate presentation of one of the type four scene signals. As embodied herein, such logic means comprises computer 12 programmed in the manner specified.
It will be apparent to those skilled in the art that modifications and variations can be made in the interactive video drama apparatus and methods of this invention. The invention and its broader aspects is not limited to the specific details, representative methods and apparatus, and illustrative example shown and described. Departure may be made from such details without departing from the spirit or the scope of the general inventive concept. ##SPC1##
Claims
I claim:
1. An interactive audiovisual system, comprising: means for storing a plurality of prerecorded audio and video scene signals representative of dramatic scenes; means for designating specific ones of said scene signals as type one, type two, type three, or type four scene signals respectively representing scenes for which relevant thoughts exist, scenes for which no relevant thoughts exists, scenes depicting relevant thoughts, and scenes depicting non-relevant thoughts, for associating each of type three scene signals with one of said type one scene signals, and for associating each of said type four scene signals with all of said type two scene signals; voice recognition means for receiving a voice signal and for generating interrupt signals representative of the meaning of words contained in said voice signals; audiovisual output means for providing audible and visual outputs representative of said scene signals; logic means for selectively presenting specific ones of said scene signals to said audiovisual output means in response to interrupt signals, for responding to said interrupt signals during presentation of said type one scene signals to alter presentation of the type one scene signal being presented, initiate presentation of one of said associated type three scene signals, and resume presentation of the type one scene signal being presented, and for responding to said interrupt signals during presentation of said type two scene signals to alter presentation of the type two scene signal being presented, initiate presentation of one of said type four scene signals, and resume presentation of the type two scene signal being presented.
2. A system as recited in claim 1, wherein said means for storing prerecorded audio and video scene signals comprises a read-only device and said logic means comprises a programmable computer.
3. A system as recited in claim 2 wherein said read-only device comprises a programmable videodisc player.
4. A system as recited in claim 3 comprising a read-write memory device for storing said type three and type four scene signals including audio components of dramatic scenes, and wherein said logic means is responsive to said interrupt signals to simultaneously supply to said audiovisual output means said audio components of said type three and type four scene signals from said read-write memory device and video components of said type one and type two scene signals from said videodisc player.
5. A system as recited in claim 4 wherein said type three and type four scene signals stored in said read-write memory device comprise digital signals representative of human speech and said logic means comprises means for generating electrical signals convertable by said audiovisual output means to human speech.
6. An interactive audiovisual system, comprising: voice recognition means for receiving voice input signals and for generating interrupt signals representative of the meaning of words contained in said voice input signals; means for storing a plurality of prerecorded audio and video scene signals representative of dramatic scenes; means for designating a plurality of said dramatic scenes as a scene group, for designating specified ones of said interrupt signals as an interrupt signal group, and for associating scenes of said scene group with interrupt signals of said interrupt signal group; audiovisual output means for providing audible and visual outputs representative of said scene signals; and logic means for selectively presenting specific ones of said scene signals to said audiovisual output means in response to said interrupt signals, for supplying only one of said scenes of said scene group to said audiovisual output means upon occurrence of one of said interrupt signals of said interrupt signal group, for recording which of said dramatic scenes have been supplied to said audiovisual output means, and for ensuring that repetitive generation of any of said interrupt signals of said interrupt signal group will result in no more than a single display of a scene of said scene group.
7. An interactive computer system, comprising: means for receiving input signal and for generating interrupt signals representative of said input signals; means for storing a plurality of prerecorded output signals and for storing a subset of said output signals as decision point outcome signals; means for storing a plurality of addresses each associated with one of said output signals and for storing a subset of said addresses as decision point addresses each associated with a predetermined plurality of said decision point outcome signals; audiovisual output means for displaying said prerecorded output signals; and logic means for selectively presenting specific ones of said stored output signals to said audiovisual output means in response to said input signals, said logic means comprising decision point means for selecting a specified one of said decision point outcome signals based on predetermined probability factors, whereby the sequence of said output signals presented to said audiovisual output means varies in accordance with said predetermined probability factors.
8. A system as recited in claim 7 wherein said logic means further comprises means for interpreting specified ones of said input signals as influence factors and for altering said predetermined probability factors according to said influence factors.
9. A system as recited in claim 8, wherein said means for interpreting specified ones of said input signals as influence factors and for altering said predetermined probability factors comprises means for storing a plurality of probability ranges each associated with a specified combination of said influence factors, means for storing a record of the occurrence of said influence factors contained in said specified combinations, and means for replacing said predetermined probability factors with revised discrete probability values calculated from said stored record of the occurrence of said input signals contained in said specified combinations and from said probability ranges.
10. A system as recited in claim 9, wherein said means for replacing said probability ranges comprises: means for associating random discrete probability component values to each of said probability ranges, said random discrete probability component values each being within the associated probability range; and means for calculating said probability values for each of said combinations of influence factors according to said random discrete probability component values.
11. A system as recited in claim 9 wherein said logic means comprises means for supplying said probability values to said audiovisual output means in response to a specified interrupt signal.
12. A method for providing an interactive drama, comprising the steps of: storing a plurality of prerecorded scene signals representative of dramatic scenes; designating specific ones of said scene signals as type one, type two, type three, or type four scene signals respectively representing scenes for which relevant thoughts exist, scenes for which no relevant thoughts exist, scenes depicting relevant thoughts, and scenes depicting nonrelevant thoughts; receiving voice input signals and generating interrupt signals representative of the meaning of words in said voice input signals; associating said type three scene signals with specific ones of said type one scene signals; selectively displaying specific ones of said stored dramatic scenes in response to said interrupt signals; responding to selected ones of said interrupt signals during presentation of said type one scene signals to alter presentation of said type one scene signal being presented, initiate presentation of one of said associated type three scene signals, and resume presentation of said type one scene signal originally presented; and responding to said selected ones of said interrupt signals during presentation of said type two scene signals to alter presentation of said type two scene signal being presented, initiate presentation of one of said associated type four scene signals, and resume presentation of said type two scene signal originally presented.
13. A method as recited in claim 12, wherein said step of storing scene signals comprises the substep of storing audio and visual components of said dramatic scenes on a read-only device.
14. A method as recited in claim 13 wherein said step of storing scene signals comprises the substep of storing audio and visual components of said dramatic scenes on a read-only videodisc.
15. A method as recited in claim 14 wherein said step of storing scene signals comprises the substep of storing said type three and type four scene signals as digital signals representative of human speech.
16. A method as recited in claim 15 wherein said step of designating said scene signals comprises the additional step of designating a subset of said scene signals as type five scene signals and wherein the step of storing scene signals comprises the substep of storing a plurality of attribute signals which associate specified ones of said interrupt signals with a plurality of said type five scene signals, and wherein said step of displaying dramatic scenes in response to interrupt signals comprises the substeps of responding to said specified ones of said interrupt signals by displaying only one of said associated type five dramatic scenes, recording which of said associated type five dramatic scenes have been displayed, and responding to said specific ones of said interrupt signals by displaying one of said associated type five scene signals which has not been previously displayed, whereby none of said associated type five dramatic scenes are displayed more than once.
17. A method as recited in claim 16 wherein the step of storing said scene signals comprises the substep of contiguously storing type five scene signals associated with a specified one of said interrupt signals, whereby said associated type five scenes can be displayed in sequence without initiating a search.
18. A method as recited in claim 17 comprising the additional step of storing type six scene signals representative of type six dramatic scenes and responding to one of said specified interrupt signals after display of all of the type five scenes associated with said one of said specified interrupt signals by displaying one of said type six dramatic scenes.
19. A method for providing interactive output of prerecorded output signals, comprising the steps of: receiving input signals and generating interrupt signals representative of said input signals; storing a plurality of prerecorded output signals, storing a first subset of said output signals as decision point signals, and storing a second subset of said output signals as decision point outcome signals, each of said decision point signals having associated therewith a plurality of decision point outcome signals; selectively outputting specific ones of said stored output signals in response to said input signals; selectively outputting, after output of one of said decision point signals, one of said decision point outcome signals associated with said one of said decision point signals, the choice of outputted decision point outcome signal being based on predetermined probability factors, whereby the sequence of said output signals being outputted varies in accordance with said predetermined probability factors.
20. A method as recited in claim 19 wherein said step of selectively outputting output signals comprises the substeps of interpreting specified ones of said input signals as influence factors and altering said predetermined probability factors according to said influence factors.
21. A method as recited in claim 20, wherein said substeps of interpreting specified ones of said input signals as influence factors and altering said predetermined probability factors comprise additional substeps of storing a plurality of probability ranges each associated with a specified combination of said influence factors, storing a record of the occurrence of said influence factors contained in said specified combinations, and replacing said predetermined probability factors with revised discrete probability values calculated from said stored record of the occurrence of said input signals contained in said specified combinations and from said probability ranges.
22. A method as recited in claim 21 wherein said step of replacing probability ranges comprises the substeps of: associating random discrete probability component values with each of said probability ranges, said random discrete probability component values being within the associated probability range; and calculating said probability values for each of said combinations of influence factors according to said random discrete probability component values.
23. A system as recited in claim 21 comprising the additional step of outputting said probability values in response to a specified interrupt signal.
24. An interactive audiovisual system, comprising: means for storing a plurality of prerecorded audio and video scene signals representative of a speaking human being, said scene signals including decision point scene signals and non-decision point scene signals; voice recognition means for storing a plurality of voice signals representative of the meaning of specified words, for receiving input voice signals, and for selectively generating interrupt signals representative of the meaning of words in said received input voice signals; and means for selectively outputting specific ones of said scene signals in response to said interrupt signals to provide an audiovisual representation of a human being speaking a meaningful response to said received input voice signals, said voice recognition means being responsive to said received input voice signals and said outputting means being responsive to said interrupt signals during output of said non-decision point signals; wherein said storing means comprises means for storing cueless scene signals devoid of indicators specifying voice input signals, and wherein said outputting means is responsive to said interrupt signals during output of said cueless scene signals; wherein said outputting means comprises means for selectively outputting specific ones of said scene signals independent of scene signals previously outputted.
25. An interactive audiovisual system, comprising: means for storing a plurality of prerecorded audio and video scene signals representative of a speaking human being, said scene signals including cueless scene signals devoid of indicators specifying voice input signals; means for receiving voice input signals and for generating interrupt signals representative of the meaning of words in said voice input signals; and means for selectively outputting specific ones of said scene signals in response to said interrupt signals to provide an audiovisual representation of a human being speaking a meaningful response to said received voice input signals, said outputting means being responsive to said interrupt signals during output of said cueless scene signals; wherein said outputting means comprises means for selectively outputting specific ones of said scene signals independent of scene signals previously outputted.
26. An interactive audiovisual system, comprising: means for storing a plurality of prerecorded audio and video scene signals representative of a speaking human being; means for storing a plurality of voice signals representative of the meaning of specified words, for receiving voice input signals and for selectively generating interrupt signals representative of the meaning of words in said stored voice signals; and means for selectively outputting specific ones of said scene signals, independent of scene signals previously outputted, in response to said interrupt signals to provide an audiovisual representation of a human being speaking a meaningful response to said received voice input signals.
27. A method for providing an interactive drama, comprising the steps of: storing a plurality of prerecorded audio and video scene signals representative of a speaking human being, said scene signals including decision point scene signals and to non-decision point scene signals; receiving voice signals and generating interrupt signals representative of the meaning of words in said voice signals; and selectively outputting specific ones of said scene signals in response to said interrupt signals to provide an audiovisual representation of a human being speaking a meaningful response to said received voice signals; said receiving and generating step including the substeps of receiving voice signals and generating interrupt signals during output of said non-decision point scene signals; wherein said storing step includes the substep of storing cueless scene signals devoid of indicators specifying voice input signals, and said receiving and generating step includes the substep of receiving voice input signals and generating interrupt signals during output of said cueless scene signals; and wherein said outputting step includes the substep of selectively outputting specific ones of said scene signals independently of previously outputted scene signals.
28. A method for providing an interactive drama, comprising the steps of: storing a plurality of prerecorded audio and video scene signals representative of a speaking human being including cueless scene signals devoid of indicators specifying voice input signals; receiving voice signals and generating interrupt signals representative of the meaning of words and said voice signals; and selectively outputting specific ones of said scene signals and response to said interrupt signals to provide an audiovisual representation of a human being speaking a meaningful response to said received voice signals; said receiving and generating step including the substep of receiving voice signals and generating interrupt signals during the output of said cueless scene signals; and wherein said outputting step includes the substep of selectively outputting specific ones of said scene signals independently of previously outputted scene signals.
29. A method for providing an interactive drama, comprising the steps of: storing a plurality of prerecorded audio and video scene signals representative of a speaking human being; storing a plurality of voice signals representative of the meaning of specified words; receiving voice signals and selectively generating interrupt signals representative of the meaning of words in said stored voice signals; and selectively outputting specific ones of said scene signals independently of previously outputted scene signals in response to said interrupt signals to provide an audiovisual representation of a human being speaking a meaningful response to said received voice signals.
Patent Citations (12)
| Patent | Date | Inventor | Cited By |
|---|---|---|---|
| US3392239 | 1968-07-01 | Johnson | |
| US3939579 | 1976-02-01 | Andrews | |
| US4130881 | 1978-12-01 | Haessler | |
| US4170832 | 1979-10-01 | Zimmerman | |
| US4305131 | 1981-12-01 | Best | |
| US4393271 | 1983-07-01 | Fujinami | |
| US4445187 | 1984-04-01 | Best | |
| US4449198 | 1984-05-01 | Kroon | |
| US4459114 | 1984-07-01 | Barwick | |
| US4482328 | 1984-11-01 | Ferguson | |
| US4571640 | 1986-02-01 | Baer | |
| US4586905 | 1986-05-01 | Groff |
Non-Patent Literature (8)
- Gilmore J., Popular Electronics, vol. 13, No. 5, Nov. 1960, pp. 60-61 and 130-132.
- Dickson, W. Patrick et al. "A Low-Cost Multimedia Microcomputer System for Educational Research and Development," Educational Technology (Aug. 1984), pp. 20-22.
- The Use of Information Technologies for Education in Science, Math and Computers, An Agenda for Research, Educational Technology Center, Cambridge, Mass. (Mar. 1984).
- Friedman, Edward A. "Machine-Mediated Instruction for Work-Force Training and Education," The Information Society (1984), vol. 2, Nos. 3/4, pp. 269-320.
- Raymont, Patrick G. "Towards Fifth Generation Training Systems," Proceedings of the IFIP WG 3.4 Working Conference on The Impact of Informatics on Vocational and Continuing Educationan (May 1984).
- Raymont, Patrick "Intelligent Interactive Instructional Systems," Microprocessing and Microprogramming (Dec. 1984), 14: 267-272.
- Best, Robert M., "Movies That Talk Back," IEEE Transactions on Consumer Electronics, vol. CE-26, Aug. 1980.
- Dickson, W. Patrick, "Experimental Software Project: Final Report," Wisconsin Center for Educational Research, University of Wisconsin, Jul. 1986.