Systems and methods for correction of text from different input types, sources, and contexts

• US 8,209,183 B1
• Filed: 07/07/2011
• Issued: 06/26/2012
• Est. Priority Date: 07/07/2011
• Status: Expired due to Fees

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Patel, Nirmal, Starner, Thad, Weaver, Josh
• Primary Examiner(s): ALBERTALLI, BRIAN LOUIS
• Application Number:
  US13/177,821
• Time in Patent Office: 355 Days
• Field of Search: None
• US Class Current: 704/270
• CPC Class Codes: G06F 40/232 Orthographic correction, e.g. spell checking or vowelisation
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/262 using context analysis, e.g. lexical, syntactic or semantic context

Abstract

Methods and apparatus related to correcting errors are disclosed. Inputs having a plurality of input types can be received at a wearable computing device. A text string corresponding to the inputs can be generated using the wearable computing device. The text string can include a plurality of segments, where each segment can be associated with an input type. For a given segment of the text string, one or more corrected segments can be generated by applying an error-correction filter configured to correct errors based on an input type associated with the given segment and a location-sensitive context. At least one of the corrected segments can be displayed using the wearable computing device. A corrected segment can be selected using the wearable computing device. A corrected text string including the selected corrected segment can be displayed using the wearable computing device.

Claims

1. A method, comprising:

receiving, at a wearable computing device, inputs having a plurality of input types;

generating a text string corresponding to the inputs using the wearable computing device, the text string comprising a plurality of segments, each segment associated with an input type of the plurality of input types;

displaying the text string using the wearable computing device by at least;

associating a first color with a first input type of the plurality of input types,

associating a second color with a second input type of the plurality of input types, wherein the first input type differs from the second input type, and wherein the first color differs from the second color,

displaying each segment of the text string associated with the first input type using the first color, and

displaying each segment of the text string associated with the second input type using the second color;

for a given segment of the text string, generating one or more corrected segments by applying an error-correction filter to the given segment using the wearable computing device, wherein the error-correction filter is configured to correct errors based on an input type associated with the given segment and on a location-sensitive context;

displaying at least the one or more corrected segments using the wearable computing device, wherein displaying the at least the one or more corrected segments comprise displaying the at least the one or more corrected segments associated with the first input type using the first color;

selecting a selected corrected segment of the one or more corrected segments using the wearable computing device; and

displaying a corrected text string comprising the selected corrected segment using the wearable computing device.

2. The method of claim 1, wherein the plurality of input types comprise speech, text entry, handwriting, and optical character recognition (OCR).

3. The method of claim 1, wherein the error-correction filter is further configured to correct errors based on a given input source.

4. The method of claim 3, wherein the input type associated with the segment is speech, and the given input source is a speaker of the speech.

5. The method of claim 1, wherein the location-sensitive context comprises information about at least one of news, current topics, and trends.

6. The method of claim 1, wherein the location-sensitive context comprises information about one or more entities proximate to a location of the wearable computing device.

7. The method of claim 6, wherein at least some of the one or more entities are people.

8. The method of claim 7, wherein the information about the one or more entities comprises information about one or more speech patterns of the people.

9. The method of claim 8, wherein displaying the least the one or more corrected segments comprises:

identifying a person of the one or more entities associated with a corrected segment of the one or more corrected segments;

updating the corrected segment based on the speech pattern of the identified person.

10. The method of claim 6, wherein the information about the one or more entities comprises identification information about the one or more entities.

11. The method of claim 10, wherein displaying the least the one or more corrected segments comprises:

identifying an entity of the one or more entities associated with a corrected segment of the one or more corrected segments; and

updating the corrected segment based on the identification information about the entity.

12. The method of claim 11, wherein updating the corrected segment based on the identification information about the entity comprises:

inserting at least part of the identification information into the corrected segment.

13. The method of claim 1, wherein the location-sensitive context comprises information about one or more speech patterns associated with a location of the wearable computing device.

14. The method of claim 13, wherein displaying the least the one or more corrected segments comprises:

identifying a location associated with a corrected segment of the one or more corrected segments;

updating the corrected segment based on a speech pattern of the one or more speech patterns associated with the location.

15. The method of claim 13, wherein the location of the wearable computing device is associated with a building, and wherein the one or more speech patterns associated with the location comprise one or more speech patterns associated with an activity associated with the building.

16. The method of claim 1, wherein the location-sensitive context is further associated with one or more documents, and

wherein displaying the least the one or more corrected segments comprises:

identifying a document of the one or more documents associated with a corrected segment of the one or more corrected segments; and

updating the corrected segment based on information in the identified document.

17. The method of claim 1, wherein the location-sensitive context is further associated with a time of a year,

wherein the location-sensitive context comprises a phrase associated with the time of the year, and

wherein displaying the least the one or more corrected segments comprises:

determining that the time of the year is associated with a corrected segment of the one or more corrected segments,

determining that the phrase associated with the time of the year is also associated with the corrected segment, and

updating the corrected segment based on the phrase associated with the time of the year.

18. A wearable computing device, comprising:

a display;

a plurality of input devices, each input device associated with an input type;

a processor; and

memory having one or more instructions that, in response to execution by the processor, causes the wearable computing device to perform functions comprising;

receiving inputs from at least some of the plurality of input devices;

generating a text string corresponding to the inputs, the text string comprising a plurality of segments, each segment associated with an input type of the plurality of input types;

displaying the text string using the display by at least;

associating a first color with a first input type of the plurality of input types,

associating a second color with a second input type of the plurality of input types, wherein the first input type differs from the second input type, and wherein the first color differs from the second color,

displaying each segment of the text string associated with the first input type using the first color, and

displaying each segment of the text string associated with the second input type using the second color;

for a given segment of the text string, generating one or more corrected segments by applying an error-correction filter to the given segment using the wearable computing device, wherein the error-correction filter is configured to correct errors based on an input type associated with the given segment and on a location-sensitive context;

displaying at least the one or more corrected segments using the display, wherein displaying the at least the one or more corrected segmented comprises displaying at least the one or more corrected segments associated with the first input type using the first color;

selecting a selected corrected segment of the one or more corrected segments; and

displaying a corrected text string comprising the selected corrected segment using the display.

19. An article of manufacture including a computer-readable storage medium having instructions stored thereon that, in response to execution by a processor, cause the processor to perform operations, the instructions comprising:

instructions for receiving inputs having a plurality of input types;

instructions for generating a text string corresponding to the inputs, the text string comprising a plurality of segments, each segment associated with an input type of the plurality of input types;

instructions for displaying the text string comprising;

instructions for associating a first color with a first input type of the plurality of input types,

instructions for associating a second color with a second input type of the plurality of input types, wherein the first input type differs from the second input type, and wherein the first color differs from the second color,

instructions for displaying each segment of the text string associated with the first input type using the first color, and

instructions for displaying each segment of the text string associated with the second input type using the second color;

instructions for generating, for a given segment of the text string, one or more corrected segments by applying an error-correction filter to the given segment, wherein the error-correction filter is configured to correct errors based on an input type associated with the given segment and on a location-sensitive context;

instructions for displaying at least the one or more corrected segments, wherein displaying the at least the one or more corrected segments comprises displaying at least the one or more corrected segments associated with the first input type using the first color;

instructions for selecting a selected corrected segment of the one or more corrected segments; and

instructions for displaying a corrected text string comprising the selected corrected segment.