WaveStitch: Flexible and Fast Conditional Time Series Generation With Diffusion Models
Summary: WaveStitch: diffusion model with dual-source conditioning on metadata and partially observed signals, using a hybrid train/inference design (metadata in training, gradient-based guidance on observations at inference). Parallel window generation with stitching preserves coherence, giving ~1.8× lower MSE and up to 166× faster sampling vs autoregressive SOTA. (summarized by gpt-5-mini on Feb 11 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Aditya Shankar (Delft University of Technology)
- 2. Lydia Chen (Delft University of Technology; University of Neuchâtel)
- 3. Arie van Deursen (Delft University of Technology)
- 4. Rihan Hai (Delft University of Technology)
BibTeX Citation
@inproceedings{shankar_sigmod26,
title = {{WaveStitch: Flexible and Fast Conditional Time Series Generation With Diffusion Models}},
author = {Shankar, Aditya and Chen, Lydia and van Deursen, Arie and Hai, Rihan},
series = {{SIGMOD} '26},
booktitle = {Proceedings of the {ACM} {SIGMOD} International Conference on Management of Data},
publisher = {Association for Computing Machinery},
doi = {10.1145/3769842},
url = {https://dl.acm.org/doi/10.1145/3769842},
year = {2026}
}Incoming Citations (Sorted by Pagerank)
Showing 0 of 0 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 7 of 7 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 1,323 | Data Cleaning: Overview and Emerging Challenges | 2016 | SIGMOD | 0.00011152602 |
| 2,129 | Data Synthesis based on Generative Adversarial Networks | 2018 | VLDB | 9.1266572e-05 |
| 3,369 | TFB: Towards Comprehensive and Fair Benchmarking of Time Series Forecasting Methods | 2024 | VLDB | 7.4706661e-05 |
| 5,158 | Forecasting Big Time Series: Old and New | 2018 | VLDB | 6.3404905e-05 |
| 5,420 | ImDiffusion: Imputed Diffusion Models for Multivariate Time Series Anomaly Detection | 2024 | VLDB | 6.2246363e-05 |
| 6,555 | Controllable Tabular Data Synthesis Using Diffusion Models | 2024 | SIGMOD | 5.8415111e-05 |
| 8,346 | TSGBench: Time Series Generation Benchmark | 2024 | VLDB | 5.4473607e-05 |
Previous
Page 1 / 1
Next
Semantically Similar Papers
| # | Overall Rank | Paper | Year | Venue |
|---|---|---|---|---|
| 1 | 10,861 | Less is More: Efficient Time Series Dataset Condensation via Two-fold Modal Matching | 2025 | VLDB |
| 2 | 8,915 | A Multi-Scale Decomposition MLP-Mixer for Time Series Analysis | 2024 | VLDB |
| 3 | 13,318 | Fully Automated Correlated Time Series Forecasting in Minutes | 2025 | VLDB |
| 4 | 5,986 | AutoTSAD: Unsupervised Holistic Anomaly Detection for Time Series Data | 2024 | VLDB |
| 5 | 6,078 | Multiple Time Series Forecasting with Dynamic Graph Modeling | 2024 | VLDB |
| 6 | 10,833 | BiST: A Lightweight and Efficient Bi-directional Model for Spatiotemporal Prediction | 2025 | VLDB |
| 7 | 8,346 | TSGBench: Time Series Generation Benchmark | 2024 | VLDB |
| 8 | 6,555 | Controllable Tabular Data Synthesis Using Diffusion Models | 2024 | SIGMOD |
| 9 | 13,320 | A Memory Guided Transformer for Time Series Forecasting | 2025 | VLDB |
| 10 | 5,420 | ImDiffusion: Imputed Diffusion Models for Multivariate Time Series Anomaly Detection | 2024 | VLDB |