Methods and apparatus for facilitating ambient content recognition using digital watermarks, and related arrangements

• US 9,905,233 B1
• Filed: 08/07/2015
• Issued: 02/27/2018
• Est. Priority Date: 08/07/2014
• Status: Active Grant

• Current Assignee: Digimarc Corporation
• Original Assignee: Digimarc Corporation
• Inventors: Thagadur Shivappa, Shankar, Rodriguez, Tony F., Eaton, Kurt M.
• Primary Examiner(s): Maung, Thomas
• Application Number:
  US14/821,435
• Time in Patent Office: 935 Days
• CPC Class Codes: G06F 21/16 Program or content traceability, e.g. by watermarking
• CPC Class Codes: G06T 1/0071 using multiple or alternating watermarks
• CPC Class Codes: G10L 19/018 Audio watermarking, i.e. embedding inaudible data in the audio signal
• CPC Class Codes: H04N 21/8358 involving watermark protecting executable software by watermarking G06F21/16; image watermarking in general G06T1/0021; watermarks inserted in still images for transmission purposes H04N1/32144; inserting watermarks during video coding H04N19/467

Abstract

The present disclosure relates generally to signal processing techniques for content signals such as audio, images and video signals. More particularly, the present disclosure relates to processing content signals to facilitate recognition of ambient content signals using digital watermarks and/or digital fingerprints.

Claims

1. A method, comprising:

obtaining a sample of ambient audio content comprising a content signal having multiple instances of a watermark message embedded therein, the watermark message conveying signal data representing a characteristic of the content signal;

utilizing one or more programmed processors for;

aggregating multiple instances of the watermark message;

decoding the watermark message from the aggregated multiple instances, thereby decoding the signal data representing the characteristic of the content signal from the watermark message;

deriving a characteristic of the sample of ambient audio content, in which the derived characteristic comprises a plurality of anchors, with each anchor of the plurality of anchors representing a location in a reference content signal corresponding to a fingerprint of the sample of ambient audio content;

modifying the derived characteristic of the sample of ambient audio content based on the decoded signal data representing the characteristic of the content signal, thereby producing a modified characteristic of the sample of ambient audio content, wherein modifying the derived characteristic comprises modifying at least a portion of the derived characteristic based on the decoded signal data representing the characteristic of the content signal conveyed by the watermark message, and wherein modifying the derived characteristic comprises performing a noise reduction process on at least a portion of the derived characteristic to remove at least one of the plurality of anchors;

accessing a data repository containing data representing a set of reference characteristics, each reference characteristic of the set of reference characteristics corresponding to one or more reference content signals; and

by reference to the modified characteristic and a reference characteristic, determining whether the sample of ambient audio content includes the reference content signal corresponding to the fingerprint of the sample of ambient audio content.

2. The method of claim 1, wherein the signal data represents a plurality of characteristics of the content signal.

3. The method of claim 1, wherein a characteristic represented by the signal data comprises an anchor within the content signal, the anchor representing a location in the content signal corresponding to a fingerprint of the ambient audio content.

4. The method of claim 3, wherein a characteristic represented by the signal data does not include the fingerprint.

5. The method of claim 1, wherein determining whether the sample corresponds to the reference content signal comprises decoding the signal data to discern a characteristic of the content signal conveyed by the watermark message.

6. The method of claim 1, wherein the derived characteristic comprises a plurality of fingerprints of the sample.

7. The method of claim 1, wherein the reference characteristic of the reference content signal comprises:

a plurality of anchors, each anchor representing a location in the reference content signal corresponding to a fingerprint of the sample; and

a plurality of fingerprints of the reference content signal, each fingerprint corresponding to an anchor.

8. The method of claim 7, further comprising decoding the signal data to discern a characteristic of the content signal conveyed by the watermark message.

9. The method of claim 1, further comprising performing a sparse vector reconstruction process on the decoded signal data to discern the characteristic of the content signal conveyed by the watermark message.

10. The method of claim 1, wherein the noise reduction process includes a de-noising process.

11. The method of claim 7, further comprising:

generating correspondences between the anchors in the modified derived characteristics and the anchors in the reference characteristics; and

determining that the sample includes a reference content signal when a plurality of corresponding anchors have a predetermined relationship.

12. The method of claim 11, wherein the predetermined relationship is present when the plurality of corresponding anchors are at least substantially linearly related.

13. The method of claim 12, wherein the predetermined relationship is present when the plurality of corresponding anchors are linearly related.

14. A method, comprising:

obtaining a sample of ambient content;

analyzing a frequency domain representation of the sample of ambient content to determine whether the sample of ambient content comprises a content signal having a watermark message embedded therein;

upon determining that the sample of ambient content comprises the content signal, determining whether the watermark message conveys signal data representing a characteristic of the content signal;

upon determining that the sample contains the watermark message, utilizing one or more processors for;

decoding the signal data from the watermark message;

deriving a characteristic of the sample of ambient content, in which the derived characteristic comprises a plurality of anchors, with each anchor of the plurality of anchors representing a location in a reference content signal corresponding to a fingerprint of the sample of ambient content; and

modifying the derived characteristic of the sample of ambient content based on the decoded signal data from the watermark message, thereby producing a modified derived characteristic of the sample of ambient content, wherein modifying the derived characteristic comprises modifying at least a portion of the derived characteristic based on the characteristic of the ambient signal conveyed by the watermark message, and wherein modifying the derived characteristic comprises performing a noise reduction process on at least a portion of the derived characteristic to remove at least one of the plurality of anchors;

accessing a data repository containing data representing a set of reference characteristics, each reference characteristic of the set of reference characteristics corresponding to one or more reference content signals; and

by reference to the modified derived characteristic and a reference characteristic, determining whether the sample of ambient content includes the reference content signal corresponding to the fingerprint of the sample of ambient content.