TimePoint: Accelerated Time Series Alignment via Self-Supervised Keypoint and Descriptor Learning

2025-05-29

AI Summary

Time-series alignment faces challenges including high computational complexity of Dynamic Time Warping (DTW), sensitivity to noise, and poor generalization. This paper proposes the first self-supervised keypoint learning framework for 1D time series: it models nonlinear temporal distortions via 1D diffeomorphisms and introduces a synthetic-data-driven joint learning paradigm for keypoints and descriptors; employs fully convolutional and wavelet convolutional networks for sparse, robust feature extraction; and performs lightweight DTW in the descriptor space. The method requires only synthetic data for training yet achieves cross-domain generalization on real-world data, accelerates alignment by several-fold over standard DTW, significantly improves alignment accuracy, and supports end-to-end fine-tuning. Code is publicly available.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Search and Optimization: Learning to Search
• Computer Vision: Diffusion Models for Vision

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Fast and scalable alignment of time series is a fundamental challenge in many domains. The standard solution, Dynamic Time Warping (DTW), struggles with poor scalability and sensitivity to noise. We introduce TimePoint, a self-supervised method that dramatically accelerates DTW-based alignment while typically improving alignment accuracy by learning keypoints and descriptors from synthetic data. Inspired by 2D keypoint detection but carefully adapted to the unique challenges of 1D signals, TimePoint leverages efficient 1D diffeomorphisms, which effectively model nonlinear time warping, to generate realistic training data. This approach, along with fully convolutional and wavelet convolutional architectures, enables the extraction of informative keypoints and descriptors. Applying DTW to these sparse representations yield major speedups and typically higher alignment accuracy than standard DTW applied to the full signals. TimePoint demonstrates strong generalization to real-world time series when trained solely on synthetic data, and further improves with fine-tuning on real data. Extensive experiments demonstrate that TimePoint consistently achieves faster and more accurate alignments than standard DTW, making it a scalable solution for time-series analysis. Our code is available at https://github.com/BGU-CS-VIL/TimePoint

Problem

• Accelerates DTW-based time series alignment via self-supervised learning
• Improves alignment accuracy by detecting keypoints and descriptors
• Generalizes to real-world data using synthetic training and fine-tuning

Innovation

• Self-supervised keypoint and descriptor learning
• Efficient 1D diffeomorphisms for synthetic data
• Fully convolutional and wavelet architectures