Method and System For Content Sampling and Identification

• US 20080154401A1
• Filed: 04/19/2005
• Published: 06/26/2008
• Est. Priority Date: 04/19/2004
• Status: Active Grant

• Current Assignee: Apple Inc.
• Original Assignee: Landmark Digital Services LLC
• Inventors: Wang, Avery Li-Chun
• Granted Patent: US 8,688,248 B2
• Time in Patent Office: Days
• US Class Current: 700/94
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 17/26 Recognition of special voice characteristics, e.g. for use in lie detectors; Recognition of animal voices
• CPC Class Codes: G10L 19/04 using predictive techniques
• CPC Class Codes: G11B 27/11 by using information not detectable on the record carrier
• CPC Class Codes: H04H 20/14 for monitoring programmes
• CPC Class Codes: H04H 60/58 of audio determination or detection of speech characteristics in general G10L25/00; speech recognition in general G10L15/00

Abstract

A method and system for content sampling (106) and identification is presented. A data stream is recorded, and samples of the stream are identified. Samples (106) can be initially taken at random for identification. Once a sample (106) is identified and segmented within the data stream, the next time to sample (106) may be calculated to be outside the time frame of the identified sample (106). Thus, the sampling period can be adaptively adjusted to be at times after identified tracks.

Claims

1. A method for segmenting a broadcast data stream comprising:

receiving a sample of the broadcast data stream;

receiving a start time of the sample;

determining a track identity of the sample;

determining a relative time offset of the sample from a beginning of the identified track; and

calculating an extrapolated start time of the identified track within the broadcast data stream using the relative time offset and the start time of the sample, wherein the extrapolated start time indicates a boundary between tracks in the broadcast data stream.

2. The method of claim 1, wherein calculating the extrapolated start time comprises:

determining a relative playback speed ratio of the identified track; and

adjusting the extrapolated start time and the extrapolated end time due to the relative playback speed.

3. The method of claim 1, further comprising determining an extrapolated end time of the sample, wherein the identified track comprises a segment of the broadcast data stream between the extrapolated start time and the extrapolated end time.

4. The method of claim 3, further comprising marking a time interval of the broadcast data stream between the extrapolated start time and the extrapolated end time as recognized, thereby extending a recognized segment within the broadcast data stream beyond the sample.

5. The method of claim 4, further comprising calculating the next time to sample the broadcast data stream based on a length of the recognized segment.

6. The method of claim 3 further comprising copying the segment of the broadcast data stream into a file.

7. The method of claim 6, further comprising:

retrieving a metadata associated with the track identity; and

associating the metadata with the sample.

8. The method of claim 7, further comprising:

organizing information of the segment of the broadcast data stream into a playlist according to the metadata;

displaying the playlist to a user;

allowing the user to select a file from the playlist; and

playing the file on an audio playback means.

9. The method of claim 3, wherein determining the extrapolated end time comprises:

obtaining a track length of the identified track; and

adding the track length to the extrapolated start time.

10. The method of claim 9, wherein obtaining the track length comprises looking up the track length based on the track identity.

11. The method of claim 1, wherein calculating the extrapolated start time of the sample comprises subtracting the relative time offset from the start time of the sample.

12. The method of claim 1, wherein determining the track identity comprises accessing a database containing track identities.

13. The method of claim 1, further comprising writing the extrapolated start time and the track identity into a file.

14. The method of claim 13, further comprising sending the extrapolated start time and the track identity to a subscriber.

15. The method of claim 1, wherein the broadcast data stream is an audio stream.

16. A method for identifying content within an audio stream comprising:

receiving samples of an audio stream;

receiving start times of each sample;

for each sample, determining a track identity and an extrapolated start time of the identified track within the audio stream;

identifying samples having substantially similar extrapolated start times; and

for samples having substantially similar extrapolated start times, verifying that the samples also have matching track identities.

17. The method of claim 16, further comprising, for each sample, determining an extrapolated end time of the identified track, wherein the identified track comprises a segment of the audio stream between the extrapolated start time and the extrapolated end time.

18. The method of claim 17, wherein determining the extrapolated end time comprises:

obtaining a track length of the identified track; and

adding the track length to the extrapolated start time.

19. The method of claim 17, further comprising marking a time interval of the audio stream between the extrapolated start time and the extrapolated end time as recognized, thereby extending a recognized audio segment within the audio stream beyond the sample.

20. The method of claim 19, further comprising calculating the next time to sample the audio stream based on a length of the recognized audio segment.

21. A method for identifying content within an audio stream comprising:

receiving an audio stream;

selecting a portion of audio from the audio stream, the portion of audio comprising a sample from the audio stream;

determining a track identity of the sample;

calculating an extrapolated start time and an extrapolated end time of the identified track within the audio stream, wherein the extrapolated start time and extrapolated end time indicate boundaries between tracks in the audio stream;

marking a segment of the audio stream between the extrapolated start time and the extrapolated end time as a recognized audio segment; and

calculating the next time to sample the audio stream based on a length of the recognized audio segment.

22. The method of claim 21, further comprising determining a confidence level indicating a degree to which a correct identity of the sample has been determined.

23. The method of claim 22, wherein if an identification attempt does not return a result, then selecting a new portion of audio for identification adjacent to the portion of audio formerly selected for identification.

24. The method of claim 21, further comprising:

selecting multiple portions of audio close in time from the audio stream;

for each portion, determining a track identity and an extrapolated start time of the identified track within the audio stream;

identifying samples having substantially similar extrapolated start times; and

for samples having substantially similar extrapolated start times, verifying that the samples also have matching track identities.

25. The method of claim 21, further comprising submitting unidentifiable segments of audio within the audio stream to a human listener to identify the segments.

26. The method of claim 21, wherein calculating the next time to sample the audio stream based on the length of the recognized audio segment comprises calculating the next time to sample to be after a time of the identified sample.

27. The method of claim 21, wherein calculating the next time to sample the audio stream based on the length of the recognized audio segment comprises adaptively adjusting a sampling interval of the audio stream.

28. The method of claim 21, wherein calculating the next time to sample the audio stream based on the length of the recognized audio segment comprises calculating a time to sample to be approximately immediately after a current audio track end time occurs.