Adaptive Multi-Trace Carving for Robust Frequency Tracking in Forensic Applications

In the field of information forensics, many emerging problems involve a critical step that estimates and tracks weak frequency components in noisy signals. It is often challenging for prior art tracking to i) achieve high accuracy under conditions, ii) detect track multiple efficiently, or iii) strike good trade-off processing delay versus resilience tracking. To address these issues, we propose Adaptive Multi-Trace Carving (AMTC), unified approach detecting one more subtle very low signal-to-noise ratio (SNR) conditions near real time. AMTC takes as input time-frequency representation system's preprocessing results (such spectrogram), identifies through iterative dynamic programming adaptive trace compensation. The proposed algorithm considers relatively energy traces sustaining over certain duration an indicator presence frequency/oscillation interest their time-varying trend. Extensive experiments using both synthetic data real-world forensic power signatures physiological monitoring reveal method outperforms representative SNR can be implemented real-time settings. empower development new technologies harness small