Arrangements for Increasing Detection Confidence

• US 20130104010A1
• Filed: 12/02/2011
• Published: 04/25/2013
• Est. Priority Date: 10/20/2011
• Status: Active Grant

• Current Assignee: Digimarc Corporation
• Original Assignee: Digimarc Corporation
• Inventors: Rhoads, Geoffrey B., Stach, John
• Granted Patent: US 9,008,192 B2
• Time in Patent Office: Days
• US Class Current: 714/799
• CPC Class Codes: G06T 1/0021 Image watermarking
• CPC Class Codes: G06T 2201/0065 Extraction of an embedded watermark; Reliable detection
• CPC Class Codes: G06V 30/248 involving plural approaches, e.g. verification by template match; Resolving confusion among similar patterns, e.g. "O" versus "Q" G06V30/242 takes precedence
• CPC Class Codes: G06V 30/2552 Combination of methods, e.g. classifiers, working on different input data, e.g. sensor fusion
• CPC Class Codes: G10L 19/018 Audio watermarking, i.e. embedding inaudible data in the audio signal

Abstract

In one embodiment, a first set of digital data (e.g., an image) is tested for the presence of a certain feature (e.g., a certain face), yielding one of two outcomes (e.g., not-present, or present). If the testing yields the first outcome, no additional testing is performed. If, however, the testing yields the second outcome, further testing is performed to further check this outcome. Such further testing is performed on a second set of digital data that is based on, but different from, the first set of data. Only if the original testing and the further testing both yield the same second outcome is it treated as a valid result. A variety of other features and arrangements are also detailed.

Claims

1. A method comprising:

receiving content data;

performing a first test on the content data to detect the presence of a first feature;

only after detecting the presence of the first feature by said first test, performing a second test to confirm the presence of the first feature;

wherein the method further includes;

corrupting the first content data using a hardware device configured to perform such act; and

performing the second test on the corrupted content data to confirm the presence of the first feature.

2. The method of claim 1 in which the first and second tests comprise the same test.

3. The method of claim 1 in which the content data comprises audio samples, and the corrupting comprises discarding or altering one or more of said samples.

4. The method of claim 1 in which the content data comprises coefficients representing audio in a frequency domain, and the corrupting comprises discarding or altering one or more of said coefficients.

5. The method of claim 1 in which the content data comprises audio data, and the corrupting comprises resampling said audio data.

6. The method of claim 1 in which the content data comprises audio data, and the. first test comprises a digital watermark decoding process.

7. The method of claim 1 in which the content data comprises audio data, and the first test comprises a pattern recognition process.

8. The method of claim 1 in which the content data comprises image data representing rows and columns of pixel data, and the corrupting comprises discarding or altering data representing one or more of said rows and/or columns.

9. The method of claim 1 in which the content data comprises coefficients representing imagery in a spatial frequency domain, and the corrupting comprises discarding or altering one or more of said coefficients.

10. The method of claim 1 in which the content data comprises image data, and the corrupting comprises virtually rotating the image data.

11. The method of claim 1 in which the content data comprises image data, and the corrupting comprises resampling the image data.

12. The method of claim 1 in which the content data comprises image data, and the first test comprises a digital watermark decoding process.

13. The method of claim 1 in which the content data comprises image data, and the first test comprises a pattern recognition process.

14. A method comprising:

(a) conducting a testing process by applying a first set of data to a testing module, the first set Of data representing image, audio or video content, the testing module outputting data indicating a testing outcome;

(b) in response to the testing process outputting data indicating a first testing outcome, conducting a re-testing process by repeating act (a) using a second set of data representing said content, the second set of data being derived from the first set;

(c) in response to the re-testing process again outputting data indicating said first testing outcome, outputting a signal indicating the first testing outcome as a final outcome.

15. The method of claim 14 in which the testing module performs an object recognition operation.

16. The method-of claim 14 in which the testing module performs a digital watermark decoding operation.

17. The method of claim 14 in which (b) comprises altering the first set of data to yield the second set of data.

18. The method of claim 17 in which the altering comprises omitting one or more elements of the first set of data to yield the second set of data.

19. The method of claim 17 in which the content comprises data representing plural blocks of image data, each block including a grouping of plural pixels, and the altering comprises changing a grouping of plural pixels in one or more of said blocks.

20. A method comprising:

receiving content data;

performing initial processing on the content data, yielding intermediate data;

performing a test to detect the presence of a feature in the content data, the test using the intermediate data; and

repeating the test using an altered version of the intermediate data to confirm the presence of said feature.

21. The method of claim 20 that includes performing said repeating act only if the test detected the presence of said feature in the content data.