Stabilizing Training With Soft Dynamic Time Warping: A Case Study For Pitch Class Estimation With Weakly Aligned Targets

Abstract

Soft dynamic time warping (SDTW) is a differentiable loss function that allows for training neural networks from weakly aligned data. Typically, SDTW is used to iteratively compute and refine soft alignments that compensate for temporal deviations between the training data and its weakly annotated targets. One major problem is that a mismatch between the estimated soft alignments and the reference alignments in the early training stage leads to incorrect parameter updates, making the overall training procedure unstable. In this paper, we investigate such stability issues by considering the task of pitch class estimation from music recordings as an illustrative case study. In particular, we introduce and discuss three conceptually different strategies (a hyperparameter scheduling, a diagonal prior, and a sequence unfolding strategy) with the objective of stabilizing intermediate soft alignment results. Finally, we report on experiments that demonstrate the effectiveness of the strategie

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyzeitler2023stabilizing

Related papers

• Soft Dynamic Time Warping For Multi-pitch Estimation And Beyond (2023)6.34
• Unsupervised Harmonic Parameter Estimation Using Differentiable DSP And Spectral Optimal Transport (2023)5.84
• Singing Voice Correction Using Canonical Time Warping (2017)5.84
• Feature Trajectory Dynamic Time Warping For Clustering Of Speech Segments (2018)7.50
• PESTO: Pitch Estimation With Self-supervised Transposition-equivariant Objective (2023)6.34
• Deep-learning Architectures For Multi-pitch Estimation: Towards Reliable Evaluation (2022)0.00
• Text-independent Speaker Verification Based On Deep Neural Networks And Segmental Dynamic Time Warping (2018)3.58
• PESTO: Real-time Pitch Estimation With Self-supervised Transposition-equivariant Objective (2025)6.34