Computer-implemented methods for synthetic accuracy measurement of a content recognition system

1. A computer-implemented method comprising:

generating, by a provider network, a reference fingerprint for a secondary content media file;

generating, by the provider network, a synthetic fingerprint for a transformed version of the secondary content media file;

inserting, by the provider network, the synthetic fingerprint into a stream of fingerprints of a plurality of media files;

comparing, by a comparison service of the provider network, the stream of fingerprints including the synthetic fingerprint to the reference fingerprint to generate an indication of a match between the synthetic fingerprint and the reference fingerprint in the stream; and

sending the indication of the match to a storage location.

2. The computer-implemented method of claim 1 , wherein the indication is at a fingerprint level and a media file level.

3. The computer-implemented method of claim 1 , further comprising:

performing an encode of the secondary content media file, used to generate the reference fingerprint, to generate an encoded secondary content media file; and

decoding the encoded secondary content media file to generate the transformed version of the secondary content media file.

4. A computer-implemented method comprising:

generating, by a provider network, a synthetic fingerprint for a transformed version of a media file;

inserting, by the provider network, the synthetic fingerprint into a stream of fingerprints of a plurality of media files;

comparing, by a comparison service of the provider network, the stream of fingerprints including the synthetic fingerprint to a reference fingerprint for an untransformed version of the media file to generate an indication of a match between the synthetic fingerprint and the reference fingerprint in the stream; and

sending the indication of the match to a storage location.

5. The computer-implemented method of claim 4 , wherein the indication is at a fingerprint level and a media file level.

6. The computer-implemented method of claim 4 , further comprising:

performing an encode of the untransformed version of the media file to generate an encoded media file; and

decoding the encoded media file to generate the transformed version of the media file.

7. The computer-implemented method of claim 4 , further comprising generating the transformed version of the media file by simulating a transformation that the untransformed version of the media file is to experience from when the media file is submitted as a source for reference fingerprint generation until the media file is received from the provider network by a user device.

8. The computer-implemented method of claim 4 , further comprising converting the untransformed version of the media file in a first format to a second format to generate the transformed version of the media file.

9. The computer-implemented method of claim 4 , further comprising:

generating, by the provider network, a second synthetic fingerprint for a transformed version of a second media file;

inserting, by the provider network, the second synthetic fingerprint into the stream of fingerprints of the plurality of media files;

comparing, by the comparison service of the provider network, the stream of fingerprints including the second synthetic fingerprint to a second reference fingerprint for an untransformed version of the second media file to generate a second indication of a second match between the second synthetic fingerprint and the second reference fingerprint in the stream; and

sending the second indication of the second match to the storage location.

10. The computer-implemented method of claim 4 , further comprising generating, by the provider network, a metadata label of the synthetic fingerprint that indicates the synthetic fingerprint was generated by the provider network.

11. The computer-implemented method of claim 10 , wherein the generation of the indication of the match between the synthetic fingerprint and the reference fingerprint in the stream is not based on the metadata label.

12. The computer-implemented method of claim 4 , wherein the synthetic fingerprint comprises an audio fingerprint.

13. The computer-implemented method of claim 4 , further comprising generating, by the provider network, a true match rate and a false match rate for the media file in the stream based at least in part on the indication.

14. The computer-implemented method of claim 4 , further comprising modifying a matching algorithm of a content recognition service based at least in part on the indication.

15. A system comprising:

a content data store to store a reference fingerprint for an untransformed version of a media file; and

one or more electronic devices to implement a computing service, the computing service including instructions that upon execution cause the computing service to perform operations comprising:

generating a synthetic fingerprint for a transformed version of the media file,

inserting the synthetic fingerprint into a stream of fingerprints of a plurality of media files,

comparing the stream of fingerprints including the synthetic fingerprint to the reference fingerprint for the untransformed version of the media file to generate an indication of a match between the synthetic fingerprint and the reference fingerprint in the stream, and

sending the indication of the match to a storage location.

16. The system of claim 15 , wherein the indication is at a fingerprint level and a media file level.

17. The system of claim 15 , wherein the instructions upon execution cause the computing service to perform operations further comprising:

performing an encode of the untransformed version of the media file to generate an encoded media file; and

decoding the encoded media file to generate the transformed version of the media file.

18. The system of claim 15 , wherein the instructions upon execution cause the computing service to perform operations further comprising generating a metadata label of the synthetic fingerprint that indicates the synthetic fingerprint was generated by the computing service, wherein the generation of the indication of the match between the synthetic fingerprint and the reference fingerprint in the stream is not based on the metadata label.

19. The system of claim 15 , wherein the instructions upon execution cause the computing service to perform operations further comprising generating a true match rate and a false match rate for the media file in the stream based at least in part on the indication.

20. The system of claim 15 , wherein the instructions upon execution cause the computing service to perform operations further comprising modifying a matching algorithm of a content recognition service based at least in part on the indication.