CASM: Context-Aware Semi-Markov Post-Processor for Beat Tracking

2026-10-06

AI Summary

This study addresses the susceptibility of peak detection to noise interference and the reliance on global constraints in dynamic Bayesian networks for music beat tracking. To overcome these limitations, this work proposes a Context-Aware Semi-Markov (CASM) decoder as a post-processing module for neural beat trackers. The method conditions temporal modeling on local activation evidence rather than predefined global constraints, integrating deterministic safety mechanisms with ambiguity disambiguation techniques. Its primary advantage lies in significantly reducing sensitivity to calibration data without requiring backbone retraining or task-specific hyperparameter tuning. Experimental results demonstrate that CASM effectively enhances temporal continuity while preserving event-level F1 scores on the GTZAN and SMC datasets, exhibiting superior robustness compared to the DBN baseline.

Abstract

In music beat tracking, model-predicted activations must be decoded into a discrete, musically coherent event sequence. Direct peak picking closely follows local evidence but can retain spurious peaks or miss weak beats. Dynamic Bayesian networks (DBNs), a widely used structured post-processor, improve sequence consistency under predefined global tempo, meter, and transition constraints, but their behavior can depend strongly on these settings. We introduce CASM, a context-aware semi-Markov decoder that instead conditions its temporal constraint on local activation evidence. CASM also accounts for ambiguity among competing periodic interpretations, including half- and double-tempo alternatives. Deterministic safeguards prevent implausible outputs and preserve beat-downbeat consistency. Applied to fixed activations from three neural beat trackers (BeatThis, MSCNN, and TCN), CASM improves temporal continuity while preserving event-level F1 across the GTZAN and SMC datasets, without backbone retraining or dataset-specific retuning. Further analysis shows that CASM is less sensitive than the DBN baseline to the composition of the calibration data.

Problem

• Beat Tracking
• Activation Decoding
• Dynamic Bayesian Networks
• Post-Processing
• Temporal Consistency

Innovation

• Beat Tracking
• Semi-Markov Decoder
• Context-Aware Post-Processor
• Dynamic Bayesian Networks
• Music Information Retrieval