发表机构
CNRS; ENS de Lyon(法国国家科学研究中心; 里昂高等师范学院)
机构由 AI 辅助整理,请以论文原文为准。AI 中文总结
本文提出TWET模型,将时间扭曲估计转化为小波域平稳化问题,利用分层膨胀卷积和可微平稳性准则进行端到端优化,在提高变形重建精度的同时显著降低计算时间,适用于低延迟场景。
AI 中文摘要
时间扭曲估计是信号处理中的一个基本问题,在生物声学、雷达和生物医学分析中有着广泛应用。本文提出了一种可训练的时间扭曲估计模型(TWET),用于从单个观测中估计时间扭曲函数。所提出的方法将时间扭曲估计表述为小波域中的平稳化问题,并利用分层膨胀卷积架构来估计时间扭曲函数。引入了一个可微的平稳性准则用于端到端优化。TWET与现有方法进行了比较。实验结果表明,变形重建精度得到提高,同时计算时间显著减少,使该框架适用于低延迟应用。
英文摘要
Time-warping estimation is a fundamental problem in signal processing with applications in bioacoustics, radar, and biomedical analysis. This paper introduces a Time-Warping Estimation Trainable (TWET) model for estimating timewarping functions from a single observation. The proposed approach formulates time-warping estimation as a stationarization problem in the wavelet domain and leverages a hierarchical dilated convolutional architecture to estimate the time-warping functions. A differentiable stationarity criterion is introduced for end-to-end optimization. TWET is compared with existing approaches. Experimental results show improved deformation reconstruction accuracy together with significantly reduced computation time, making the framework compatible with low-latency applications.
Journal ref2026 IEEE International Workshop on Machine Learning for Signal Processing, Sep 2026, Atlanta, France