Research map / Time series / Causal and generative time series
Closed-Form Diffeomorphic Transformations for Time Series Alignment
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Replace the libcpab solver with closed-form CPA integration | DIFW · CPA transformation forward operation, CPU | speedup: x18 | Helped |
| Replace the libcpab solver with a closed-form CPA gradient | DIFW · CPA transformation backward operation, CPU | speedup: x10 | Helped |
| Replace the libcpab solver with closed-form CPA integration | DIFW · CPA transformation forward operation, GPU | speedup: x260 | Helped |
| Replace the libcpab solver with a closed-form CPA gradient | DIFW · CPA transformation backward operation, GPU | speedup: x30 | Helped |
| Use DIFW temporal transformer alignment instead of Euclidean, DBA, SoftDTW, DTAN, or ResNet-TW | TTN with DIFW · UCR archive | NCC test accuracy comparison: better or no worse in 94% of the datasets compared with Euclidean, DBA (77%), SoftDTW (69%), DTAN (76%) and ResNet-TW (7 | Helped |
| Apply scaling-and-squaring to approximate the flow (ablation) | CPA diffeomorphic transformations · Large deformations | speed | Helped |
Closest papers by meaning
- Diffeomorphic Temporal Alignment Nets (NeurIPS 2019 · similarity 0.90)
- TimePoint: Accelerated Time Series Alignment via Self-Supervised Keypoint and Descriptor Learning (ICML 2025 · similarity 0.85)
- Simulation-Free Training of Neural ODEs on Paired Data (NeurIPS 2024 · similarity 0.84)
- Trajectory Flow Matching with Applications to Clinical Time Series Modelling (NeurIPS 2024 · similarity 0.84)
- Global Convergence in Neural ODEs: Impact of Activation Functions (ICLR 2025 · similarity 0.83)
- Shifting the Paradigm: A Diffeomorphism Between Time Series Data Manifolds for Achieving Shift-Invariancy in Deep Learning (ICLR 2025 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.