MULTIPLE STEP IDENTIFICATION OF RECORDINGS

• US 20100161656A1
• Filed: 03/09/2010
• Published: 06/24/2010
• Est. Priority Date: 07/31/2001
• Status: Abandoned Application

• Current Assignee: Gracenote, Inc. (RR Donnelley & Sons Company)
• Original Assignee: Gracenote, Inc. (RR Donnelley & Sons Company)
• Inventors: Roberts, Dale T., Hyman, David, White, Stephen
• Application Number:
  US12/720,550
• Publication Number:
  US 20100161656A1
• Time in Patent Office: Days
• US Class Current: 707/769
• CPC Class Codes: G06F 16/634 Query by example, e.g. query by humming
• CPC Class Codes: G06F 16/683 using metadata automatically derived from the content
• CPC Class Codes: G06F 16/7328 Query by example, e.g. a complete video frame or video sequence graphical querying G06F16/7335
• CPC Class Codes: G06F 16/78 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 16/7834 using audio features
• CPC Class Codes: G06F 16/7844 using original textual content or text extracted from visual content or transcript of audio data

Abstract

Multiple information is extracted from an unknown recording and information associated therewith. Associated information includes the filename, if the recording is a computer file in, e.g., MP3 format, or table of contents (TOC) data, if the recording is on a removable medium, such as a compact disc. At least one and preferably several algorithmically determined fingerprints are extracted from the recording using one or more fingerprint extraction methods. The information extracted is compared with corresponding information in a database maintained for reference recordings. Identification starts with the most accurate and efficient method available, e.g., using a hash ID, a unique ID or text. Fingerprint matching is used to confirm other matches and validation is performed by comparing the duration of the unknown and a possibly matching reference recording.

Claims

1. A method, comprising:

accessing unknown waveform-related information and unknown non-waveform related information of an unknown recording, the unknown waveform-related information being algorithmically derived from the unknown recording and the unknown non-waveform related information comprising metadata stored with the unknown recording;

comparing the unknown non-waveform related information with reference non-waveform related information to identify the unknown recording; and

when the comparing of the unknown non-waveform related information does not substantially uniquely identify the unknown recording, comparing the unknown waveform-related information with reference waveform-related information to substantially uniquely identify the unknown recording.

2. The method of claim 1, wherein determining when identification of the unknown recording is successful is based on an aggregation of a confidence level associated with the comparing of the unknown non-waveform related information and a confidence level associated with the comparing of the unknown waveform-related information.

3. The method of claim 1, further comprising:

validating identification of the unknown recording by comparing an extracted length of the unknown recording with a stored length of a reference recording in a database of reference recordings.

4. The method of claim 1, wherein the unknown non-waveform related information comprises a unique identifier provided by a provider of identification services, or text associated with the unknown recording.

5. The method of claim 4, wherein the text includes at least one of a file path and a filename of the unknown recording.

6. The method of claim 4, wherein the text has been derived from an ID3 tag for the unknown recording.

7. The method of claim 1, wherein the unknown waveform-related information comprises at least two fingerprints derived from the at least one of audio or video information in the unknown recording, and wherein the comparing of the unknown waveform-related information comprises:

comparing the at least two fingerprints with at least one database of correspondingly-generated fingerprints for reference recordings, and

identifying the unknown recording when a possible identification based on each of the at least two fingerprints identifies a single reference recording with respective levels of confidence.

8. The method of claim 1, further comprising:

extracting the unknown waveform-related information comprising at least two fingerprints using different fingerprinting techniques.

9. The method of claim 1, wherein the unknown recording is stored on a removable medium and the comparing of the unknown non-waveform related information includes processing data derived from table of contents data or an ID3 tag associated with the unknown recording.

10. The method of claim 1, further comprising:

maintaining a database on at least one server under control of a provider of identification services, and

transmitting from the at least one server to client equipment, extraction instructions to extract the unknown waveform-related information.

11. The method of claim 10, wherein the extraction instructions are selected based on at least one of type of the client equipment receiving the extraction instructions, geographical location of the client equipment receiving the extraction instructions, or software operating on the client equipment receiving the extraction instructions.

12. The method of claim 10, wherein the database includes identifying information associated with a plurality of reference musical recordings.

13. The method of claim 10, wherein the database includes identifying information associated with a plurality of reference video recordings.

14. The method of claim 1, further comprising:

detecting a quality of the unknown waveform-related information as a first fingerprint;

identifying another copy of the unknown recording using the first fingerprint; and

replacing the first fingerprint with a higher quality second fingerprint when the other copy of the unknown recording produces the higher quality second fingerprint.

15. The method of claim 14, wherein the detecting is based on at least one of an encoding technique used for the unknown recording, a media type used to store the unknown recording, error correction capability of user equipment accessing the unknown recording.

16. The method of claim 14, wherein the detecting is based on a number of errors detected during extracting the first fingerprint from the unknown recording.

17. The method of claim 1, wherein the metadata comprises table of contents data, and wherein the unknown waveform-related information comprises at least one fingerprint extracted from one or more tracks of the unknown recording.

18. A system, comprising:

a storage unit to access unknown non-waveform related information of an unknown recording, the unknown non-waveform related information comprising metadata stored with the unknown recording;

an extraction unit to extract unknown waveform-related information of the unknown recording, the waveform-related information being algorithmically derived from the unknown recording; and

a processor to compare the unknown non-waveform related information with reference non-waveform related information to identify the unknown recording, and when comparing the unknown non-waveform related information does not substantially uniquely identify the unknown recording, to compare the unknown waveform-related information with reference waveform-related information to substantially uniquely identify the unknown recording.

19. The system of claim 18, wherein the extraction unit is to extract at least one fingerprint from at least one of audio or video information in the unknown recording using different fingerprinting techniques.

20. The system of claim 18, wherein determining when identification of the unknown recording is successful is based on an aggregation of respective levels of confidence associated with the comparing the unknown non-waveform related information and the comparing of the unknown waveform-related information.

21. The system of claim 18, wherein the metadata comprises table of contents data, and wherein the unknown waveform-related information comprises at least one fingerprint extracted from one or more tracks of the unknown recording.

22. A method, comprising:

accessing unknown waveform-related information and unknown non-waveform related information of an unknown recording, the unknown waveform-related information being algorithmically derived from the unknown recording and the unknown non-waveform related information comprising metadata stored with the unknown recording;

comparing the unknown waveform related information with reference waveform related information to identify the unknown recording; and

when the comparing of the unknown waveform related information does not substantially uniquely identify the unknown recording, comparing the unknown non-waveform related information with reference non-waveform related information to substantially uniquely identify the unknown recording.

23. The method of claim 22, wherein the metadata comprises table of contents data, and wherein the unknown waveform-related information comprises at least one fingerprint extracted from one or more tracks of the unknown recording.

24. The method of claim 22, wherein determining when identification of the unknown recording is successful is based on an aggregation of a confidence level associated with the comparing the unknown non-waveform related information and a confidence level associated with the comparing the unknown waveform-related information.

25. The method of claim 22, further comprising:

validating identification of the unknown recording by comparing an extracted length of the unknown recording with a stored length of a reference recording in a database of reference recordings, the reference recording associated with the reference non-waveform related information and the reference waveform-related information.