Towards Human-in-the-Loop Onset Detection: A Transfer Learning Approach for Maracatu

TL;DR

This paper presents a transfer learning approach using Temporal Convolutional Networks for onset detection in Maracatu, a complex Afro-Brazilian musical tradition, achieving near-perfect accuracy with minimal annotated data.

Contribution

It introduces a novel transfer learning strategy tailored for Maracatu, demonstrating effective cross-task adaptation and minimal data requirements for accurate onset detection.

Findings

F1 scores up to 0.998 in intra-task detection

Over 50 percentage points improvement over baseline

Effective cross-task adaptation for time-keeping instruments

Abstract

We explore transfer learning strategies for musical onset detection in the Afro-Brazilian Maracatu tradition, which features complex rhythmic patterns that challenge conventional models. We adapt two Temporal Convolutional Network architectures: one pre-trained for onset detection (intra-task) and another for beat tracking (inter-task). Using only 5-second annotated snippets per instrument, we fine-tune these models through layer-wise retraining strategies for five traditional percussion instruments. Our results demonstrate significant improvements over baseline performance, with F1 scores reaching up to 0.998 in the intra-task setting and improvements of over 50 percentage points in best-case scenarios. The cross-task adaptation proves particularly effective for time-keeping instruments, where onsets naturally align with beat positions. The optimal fine-tuning configuration varies by instrument, highlighting the importance of instrument-specific adaptation strategies. This approach addresses the challenges of underrepresented musical traditions, offering an efficient human-in-the-loop methodology that minimizes annotation effort while maximizing performance. Our findings contribute to more inclusive music information retrieval tools applicable beyond Western musical contexts.