NLP tool to dynamically create movies/animated scenes

• US 20060217979A1
• Filed: 03/22/2005
• Published: 09/28/2006
• Est. Priority Date: 03/22/2005
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Pahud, Michel, Aikawa, Takako, Schwartz, Lee A.
• Granted Patent: US 7,512,537 B2
• Time in Patent Office: Days
• US Class Current: 704/257
• CPC Class Codes: G06T 13/00 Animation
• CPC Class Codes: H04N 21/23412 for generating or manipulating the scene composition of objects, e.g. MPEG-4 objects
• CPC Class Codes: H04N 21/44012 involving rendering scenes according to scene graphs, e.g. MPEG-4 scene graphs

Abstract

The subject invention provides a unique system and method that facilitates integrating natural language input and graphics in a cooperative manner. In particular, as natural language input is entered by a user, an illustrated or animated scene can be generated to correspond to such input. The natural language input can be in sentence form. Upon detection of an end-of-sentence indicator, the input can be processed using NLP techniques and the images or templates representing at least one of the actor, action, background and/or object specified in the input can be selected and rendered. Thus, the user can nearly immediately visualize an illustration of his/her input. The input can be typed, written, or spoken—whereby speech recognition can be employed to convert the speech to text. New graphics can be created as well to allow the customization and expansion of the invention according to the user'"'"'s preferences.

Claims

1. A natural language illustration system comprising:

a language processor component that analyzes natural language input to determine a logical form of the input and translates the logical form into an output, the output comprising at least one of an actor, action, object, and background; and

an animation engine that selects at least one image, applies at least one template to the at least one image and generates an animation dynamically.

2. The system of claim 1, the language processor component is a natural language processor component that comprises a natural language processor module that processes the natural language input to determine the logical form of the input.

3. The system of claim 1, the output is in XML format.

4. The system of claim 3, the animation engine comprises an XML parser component that receives XML formatted output from the language processor component and determines which images to retrieve and render based at least in part on the XML formatted output.

5. The system of claim 3, further comprising a scene generator that arranges the images in a manner consistent with a meaning of the natural language input.

6. The system of claim 1, further comprising a graphics library that comprises a plurality of images corresponding to a multitude of at least the following:

actors, actions, backgrounds, and objects.

7. The system of claim 1, the animation is generated in at least one of 2-D or 3-D space.

8. The system of claim 1, the output further comprises at least one of size, dimension, age, mood, and color.

9. The system of claim 1, the input comprises natural language text in sentence form.

10. The system of claim 9, the animation engine generates one scene at a time after each sentence is entered.

11. The system of claim 1, the language processor component begins analysis of the input upon detection of an end-of-sentence indicator in the input.

12. The system of claim 1, further comprising an image generation component that creates and stores new images that are individually associated with a particular term as determined by a user.

13. The system of claim 1, the language processor component assigns a core term to a plurality of synonyms to mitigate needing a different image for each synonym and communicates the core term to the animation engine instead of the synonym specified in the input.

14. The system of claim 1, the animation engine renders the scene after the natural language input is received by the language processor component.

15. A method that facilitates generating illustrations in an instant manner from natural language input comprising:

receiving natural language input;

processing the natural language input to determine logical form information;

translating the logical form information into a formatted output; and

selecting and dynamically rendering at least one image associated with the formatted output to generate an animation that illustrates the natural language input.

16. The method of claim 15, wherein processing the natural language input begins after detecting an end-of-sentence indicator.

17. The method of claim 15, the logical form information comprises actor, action, background, object, and other functional roles pertaining to the natural language input.

18. The method of claim 15, further comprising viewing a series of animated scenes in order of generation in a movie-like manner.

19. The method of claim 15, further comprising creating at least one new image and adding it to a graphics library from which the images are selected based on the logical form information.

20. A natural language illustration system comprising:

means for analyzing natural language input to determine a logical form of the input and translates the logical form into a formatted output, the output comprising at least one of an actor, action, object, and background;

means for selecting at least one image;

means for applying at least one template to the at least one image; and

means for generating an animation dynamically based on the at least one image and the at least one template.