Reproduce “Local Periodicity-Based Beat Tracking for Expressive Classical Piano Music” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Ching-Yu Chiu, Meinard Müller, Matthew E. P. Davies, Alvin Wen-Yu Su, Yi-Hsuan Yang
Researchers developed Predominant Local Pulse-based Dynamic Programming (PLPDP), a new post-processing tracker that incorporates local periodicity to improve beat tracking in expressive classical piano music. This method effectively adapts to continuous tempo changes, achieving an F1-score of 0.838 on the challenging Maz-5 dataset with real audio inputs.
Key insights
- Global tempo assumptions in current post-processing trackers are a fundamental limitation for accurately tracking beats in expressive classical music, even with perfect beat activation inputs.
- Incorporating local periodicity and temporal expectations, similar to human auditory cognition, allows for more robust adaptation to continuous and significant tempo variations.
- Evaluating post-processing trackers with 'synthetic activation functions' (derived from ground truth) is crucial for isolating and understanding their intrinsic performance limitations.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.