System and method for communicating media signals
View Patent ↗A media streaming system for streaming media signals is provided. The media streaming system takes a library of separate and distinct CODECs that are provided as a searchable CODEC library and used in determining specific characteristics in the media signal to identify similar sections of the signal. The media streaming system uses a computer implemented intelligence system, such as an artificial intelligence mechanism to learn and capture the unique characteristics of a sinal as the signal is being streamed. The media streaming system also compresses and decompresses the media signal as the signals are streamed from a source media to a destination device.
1. A method comprising:
obtaining a media signal to be communicated to a destination agent, the media signal being separated into a plurality of segments each comprising a number of temporally adjacent frames; and
repeating for each of the plurality of segments:
testing a plurality of different CODECs on the segment to determine how each CODEC encodes the segment in terms of quality and compression level;
automatically selecting the CODEC that produces the highest quality encoded output for the segment according to a set of criteria without exceeding a bandwidth constraint;
delivering the segment encoded using the selected CODEC to the destination agent; and
reporting to the destination agent which CODEC was used to encode the segment;
wherein at least two segments are encoded using different CODECs.
2. The method of claim 1 , further comprising storing an association between one or more identified characteristics of a segment and the selected CODEC.
3. The method of claim 2 , further comprising:
in response to a subsequent segment of the media signal being found to have the same one or more identified characteristics, automatically selecting the CODEC from the stored association to encode the subsequent segment.
4. The method of claim 3 , wherein the CODEC is automatically selected from the stored association by an artificial intelligence (AI) system.
5. The method of claim 4 , wherein the AI system comprises a neural network.
6. The method of claim 2 , wherein the characteristics of the segment are selected from the group consisting of temporal characteristics, spatial characteristics, and logical characteristics.
7. The method of claim 1 , wherein testing further comprises:
storing a baseline snapshot of the segment; and
for each CODEC to be tested:
encoding the segment at or below the bandwidth constraint using one of the CODECs;
decoding the segment using the same CODEC; and
comparing the quality of the decoded segment with the baseline snapshot according to the set of criteria.
8. The method of claim 7 , wherein comparing further comprises comparing the quality according to a Peak Signal-to-Noise Ratio (PSNR).
9. The method of claim 1 , further comprising adjusting the bandwidth constraint based on constraints of at least one of the destination agent and a transmission channel to the destination agent.
10. The method of claim 1 , wherein the CODECs are selected from the group consisting of block CODECs, fractal CODECs, and wavelet CODECs.
11. The method of claim 1 , wherein delivering further comprises transmitting the encoded segment to the destination agent through a network; and wherein reporting comprises sending an indication of which CODEC was used to encode the segment through the network to the destination agent.
12. The method of claim 1 , wherein delivering further comprises storing the encoded scene on a storage medium; and wherein reporting comprises storing an indication of which CODEC was used to encode the scene on the storage medium.
13. A system comprising:
an input module to obtain a media signal to be communicated to a destination agent, the media signal being separated into a plurality of segments each comprising a number of temporally adjacent frames;
a selection module to test a plurality of different CODECs on each of the plurality of segments to determine how each CODEC encodes each segment in terms of quality and compression level, wherein the selection module is further to select the CODEC that produces the highest quality encoded output for each segment according to a set of criteria without exceeding a bandwidth constraint;
an output module to deliver each segment encoded using a respective selected CODEC to the destination agent and report to the destination agent which CODEC was used to encode each segment.
14. The system of claim 13 , wherein the selection module tests the plurality of CODECs on a segment by storing a baseline snapshot of the segment and, for each CODEC to be tested, encoding the segment at or below the bandwidth constraint using one of the CODECs, decoding the segment using the same CODEC, and comparing the quality of the decoded segment with the baseline snapshot according to the set of criteria.
15. The system of claim 14 , wherein the quality is compared according to a Peak Signal-to-Noise Ratio (PSNR).
16. The system of claim 13 , wherein the selection module is to store an association between one or more identified characteristics of a segment and the selected CODEC.
17. The system of claim 16 , wherein the selection module, in response to a subsequent segment of the media signal being found to have the same one or more identified characteristics, is to automatically select the CODEC from the stored association to encode the subsequent segment.
18. The system of claim 17 , wherein the CODEC is automatically selected from the stored association by an artificial intelligence (AI) system.
19. The system of claim 18 , wherein the AI system comprises a neural network.
20. The system of claim 16 , wherein the characteristics of the scene are selected from the group consisting of temporal characteristics, spatial characteristics, and logical characteristics.
21. The system of claim 13 , wherein the selection module is to adjust the bandwidth constraint in response to constraints of at least one of the destination agent and a transmission channel to the destination agent.
22. The system of claim 13 , wherein the CODECs are selected from the group consisting of block CODECs, fractal CODECs, and wavelet CODECs.
23. A system comprising:
means for obtaining a media signal to be communicated to a destination agent, the media signal being separated into a plurality of segments each comprising a number of temporally adjacent frames;
means for testing a plurality of different CODECs on each of the plurality of segments to determine how each CODEC encodes the segment in terms of quality and compression level;
means for selecting the CODEC that produces the highest quality encoded output for each segment according to a set of criteria without exceeding a bandwidth constraint;
means for delivering each segment encoded using a respective selected CODEC to the destination agent and report to the destination agent which CODEC was used to encode each segment.
24. The system of claim 23 , wherein the testing means tests the plurality of CODECs on a segment by storing a baseline snapshot of the segment and, for each CODEC to be tested, encoding the segment at or below the bandwidth constraint using one of the CODECs, decoding the segment using the same CODEC, and comparing the quality of the decoded segment with the baseline snapshot according to the set of criteria.
25. The system of claim 24 , wherein the quality is compared according to a Peak Signal-to-Noise Ratio (PSNR).
26. The system of claim 23 , wherein the selection means stores an association between one or more identified characteristics of the segment on which the CODECs were tested and the selected CODEC.
27. The system claim 26 , wherein the selection means, in response to a subsequent segment of the media signal being found to have the same one or more identified characteristics, automatically selects the CODEC from the stored association to encode the subsequent segment.
28. The system of claim 27 , wherein the selection means comprises an artificial intelligence (AI) system.
29. The system of claim 28 , wherein the AI system comprises a neural network.
30. The system of claim 26 , wherein the characteristics of the scene are selected from the group consisting of temporal characteristics, spatial characteristics, and logical characteristics.
31. The system of claim 23 , wherein the selection means is to adjust the bandwidth constraint in response to constraints of at least one of the destination agent and a transmission channel to the destination agent.
32. The system of claim 23 , wherein the CODECs are selected from the group consisting of block CODECs, fractal CODECs, and wavelet CODECs.
33. A method comprising:
obtaining a media signal to be communicated to a destination agent, the media signal being separated into a plurality of segments each comprising a number of temporally adjacent frames; and
repeating for each of the plurality of segments:
simultaneously testing a plurality of different CODECs on the segment to determine how each CODEC encodes the segment in terms of quality and compression level;
automatically selecting the CODEC that produces the highest quality encoded output for the segment according to a set of criteria without exceeding a bandwidth constraint;
delivering the segment encoded using the selected CODEC to the destination agent; and
reporting to the destination agent which CODEC was used to encode the segment.
34. The method of claim 33 , wherein the CODECs are simultaneously tested on the segment using a plurality of processors operating in parallel.
35. The method of claim 33 , wherein testing further comprises:
storing a baseline snapshot of the segment; and
for each CODEC to be tested:
encoding the segment at or below the bandwidth constraint using one of the CODECs;
decoding the segment using the same CODEC; and
comparing the quality of the decoded segment with the baseline snapshot according to the set of criteria.