Methods and apparatuses for implementing adaptive multi-trace carving to track signal traces
Systems, methods, apparatuses, and computer program products for tracking weak signal traces under severe noise and/or distortions. A method may include tracking at least one candidate frequency trace from a time-frequency representation of a signal. The method may also include identifying a frequency trace of the signal based on tracking results. In addition, the method may include outputting an estimated frequency vector related to the frequency trace. Further, the tracking may be performed under a noisy condition environment.
1. A method of authenticating frequency traces, comprising:
iteratively tracking at least one candidate frequency trace from a time-frequency representation of a signal;
identifying all frequency traces of the signal via a multiple frequency traces tracking and detection procedure, wherein the multiple frequency traces tracking and detection procedure comprises
identifying at least one frequency trace comprising diffused energy of the signal within a predefined frequency range based on tracking results, and
determining an estimated frequency vector related to the at least one frequency trace by implementing an energy function, wherein the energy function comprises the time-frequency representation of the signal to be analyzed, and a regularization parameter that enforces smooth tracking results;
outputting the estimated frequency vector related to the at least one frequency trace;
reducing interference in subsequent iterations of tracking the at least one candidate frequency trace by attenuating, at an end of each iteration, diffused energy around the at least one frequency trace indicated by the estimated frequency vector; and
authenticating the identified frequency traces by validating all the identified and attenuated frequency traces against a reference source,
wherein the tracking is performed under a noisy condition environment.
2. The method according to claim 1 , wherein the tracking and outputting are performed via a convolutional neural network or the multiple frequency traces tracking and detection procedure.
3. The method according to claim 2 , wherein the convolutional neural network comprises a plurality of convolutional layers, a plurality of pooling layers, and at least one fully connected layer.
4. The method according to claim 3 , wherein each of the plurality of convolutional layers comprises a different filter size in frequency and in time.
5. The method according to claim 3 , wherein the plurality of pooling layers comprises a combination of maxpooling layers and an average pooling layer.
6. The method according to claim 3 , wherein the method further comprises applying a rectified linear units activation function after each of the plurality of convolutional layers and the at least one fully connected layer.
7. The method according to claim 2 , wherein the multiple frequency traces tracking and detection procedure comprises an iterative dynamic programming application and an adaptive trace compensation application.
8. The method according to claim 7 , wherein the iterative dynamic programming application comprises iteratively computing an accumulated regularized maximum energy map of the signal column by column for all entries.
9. The method according to claim 7 , wherein the iterative dynamic programming application further comprises backtracking from a maximum entry of a last column of the accumulated regularized maximum energy map.
10. The method according to claim 7 , wherein the iterative dynamic programming application further comprises determining, by a relative energy ratio, a presence of a frequency component within a time window of the signal.
11. The method according to claim 2 , wherein the multiple frequency traces tracking and detection procedure comprises an adaptive multi-trace carving application.
12. The method according to claim 11 , wherein the adaptive multi-trace carving application comprises iteratively running an iterative dynamic programming application, and attenuating diffused energy around the frequency component.
13. An apparatus, comprising:
at least one processor; and
at least one memory comprising computer program code,
the at least one memory and the computer program code are configured, with the at least one processor to cause the apparatus at least to
iteratively track at least one candidate frequency trace from a time-frequency representation of a signal;
identify all frequency traces of the signal via a multiple frequency traces tracking and detection procedure, wherein the multiple frequency traces tracking and detection procedure comprises
identification of at least one frequency trace comprising diffused energy of the signal within a predefined frequency range based on tracking results, and
determination of an estimated frequency vector related to the at least one frequency trace by implementing an energy function, wherein the energy function comprises the time-frequency representation of the signal to be analyzed, and a regularization parameter that enforces smooth tracking results;
output the estimated frequency vector related to the at least one frequency trace;
reduce interference in subsequent iterations of tracking the at least one candidate frequency trace by attenuating, at an end of each iteration, diffused energy around the at least one frequency trace indicated by the estimated frequency vector; and
authenticate the identified frequency traces by validating all the identified and attenuated frequency traces against a reference source,
wherein the tracking is performed under a noisy condition environment.
14. A computer program embodied on a non-transitory computer-readable medium, said computer program comprising program instructions stored thereon for performing a method, the method comprising:
iteratively tracking at least one candidate frequency trace from a time-frequency representation of signal;
identifying all frequency traces of the signal via a multiple frequency traces tracking and detection procedure, wherein the multiple frequency traces tracking and detection procedure comprises
identifying at least one frequency trace comprising diffused energy of the signal within a predefined frequency range based on tracking results, and
determining an estimated frequency vector related to the at least one frequency trace by implementing an energy function, wherein the energy function comprises the time-frequency representation of the signal to be analyzed, and a regularization parameter that enforces smooth tracking results;
outputting the estimated frequency vector related to the at least one frequency trace;
reducing interference in subsequent iterations of tracking the at least one candidate frequency trace by attenuating, at an end of each iteration, diffused energy around the at least one frequency trace indicated by the estimated frequency vector; and
authenticating the identified frequency traces by validating all the identified and attenuated frequency traces against a reference source,
wherein the tracking is performed under a noisy condition environment.