Audio content recognition method and device

1. An audio contents recognition method comprising:

receiving an audio signal;

obtaining audio fingerprints (AFPs) based on a spectral shape of the received audio signal;

generating hash codes for the obtained audio fingerprints;

transmitting a matching query about a match between the generated hash codes and hash codes stored in a database; and

receiving a contents recognition result of the audio signal in response to the transmitting,

wherein the generating of the hash codes comprises determining a frame interval of the obtained audio fingerprints based on a discrete cosine transform (DCT) coefficient difference change of reference frames among a plurality of frames of the obtained audio fingerprints and generating the hash codes for the obtained audio fingerprints based on the determined frame interval.

2. The audio contents recognition method of claim 1 , wherein the audio fingerprints are determined based on a frequency domain spectral shape of the received audio signal.

3. The audio contents recognition method of claim 2 , wherein the frame interval is generated based on a spectral size difference between adjacent frames of the obtained audio fingerprints.

4. The audio contents recognition method of claim 1 , wherein the generating of the hash codes comprises applying a weight determined based on frequency domain energy of the obtained audio fingerprints.

5. The audio contents recognition method of claim 1 , wherein the transmitting of the matching query comprises determining hash codes to transmit a matching query and a transmission priority of the hash codes to transmit the matching query among the generated hash codes based on a number of bit variations between hash codes corresponding to frames adjacent to each other.

6. The audio contents recognition method of claim 1 , wherein the contents recognition result is determined based on contents identifications (IDs) of the hash codes that transmitted the matching query and a frame concentration measure (FCM) of a frame domain.

7. The audio contents recognition method of claim 1 , wherein the audio signal comprises at least one of channel audio and object audio.

8. The audio contents recognition method of claim 1 , further comprising:

analyzing an audio scene feature of the received audio signal; and

setting a section to obtain an audio fingerprint based on the audio scene feature,

wherein the obtaining of the audio fingerprints comprises obtaining an audio fingerprint for the section.

9. The audio contents recognition method of claim 1 , further comprising:

receiving an audio contents recognition command and a matching query transmission command,

wherein the obtaining of the audio fingerprints comprises obtaining an audio fingerprint for a section from a time when the audio contents recognition command is received to a time when the matching query transmission command is received.

10. The audio contents recognition method of claim 1 , wherein the generating of the hash codes comprises, if audio fingerprints having the same value are present among the obtained audio fingerprints, deleting the audio fingerprints having the same value except for one.

11. An audio contents recognition method comprising:

receiving an audio signal;

obtaining audio fingerprints (AFPs) of the received audio signal;

generating hash codes for the obtained audio fingerprints;

matching the generated hash codes and hash codes stored in a database; and

recognizing contents of the audio signal based on a result of the matching,

wherein the generating of the hash codes comprises determining a frame interval of the obtained audio fingerprints based on a discrete cosine transform (DCT) coefficient difference change of reference frames among a plurality of frames of the obtained audio fingerprints and generating the hash codes for the obtained audio fingerprints based on the determined frame interval.

12. An audio contents recognition device comprising:

a transceiver configured to receive an audio signal;

a processor; and

a memory storing instructions executable by the processor,

wherein the processor is configured to:

obtain audio fingerprints (AFPs) of the received audio signal,

generate hash codes for the obtained audio fingerprints, transmit a matching query about a match between the generated hash codes and hash codes stored in a database, and receive a contents recognition result of the audio signal in response to the transmitting,

wherein the processor is further configured to determine a frame interval of the obtained audio fingerprints based on a discrete cosine transform (DCT) coefficient difference change of reference frames among a plurality of frames of the obtained audio fingerprints and to generate the hash codes for the obtained audio fingerprints based on the determined frame interval.

13. An audio contents recognition device comprising:

a transceiver configured to receive an audio signal;

a processor; and

a memory storing instructions executable by the processor,

wherein the processor is configured to:

obtain audio fingerprints (AFPs) of the received audio signal,

generate hash codes for the obtained audio fingerprints, and

match the generated hash codes and hash codes stored in a database and recognize contents of the audio signal based on a result of the matching,

wherein the processor is further configured to determine a frame interval of the obtained audio fingerprints based on a discrete cosine transform (DCT) coefficient difference change of reference frames among a plurality of frames of the obtained audio fingerprints and to generate the hash codes for the obtained audio fingerprints based on the determined frame interval.

14. A computer-readable recording medium having recorded thereon a computer program for implementing the method of claim 1 .