Research map / Time series / Causal and generative time series
CaT-Diff: Cascaded Text-enhanced Diffusion Model for Time-Series Imputation
Changjian Xu, Yong Wang, Ruizheng Huang, Zhicheng Zhang, Wen Yin, Kexin Li
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add Hierarchical Semantic Filter (ablation) | CaT-Diff · Time-MMD Energy / Public Health | MSE: Energy 0.161 -> 0.135; Public Health 20.234 -> 16.763 | Improved |
| add Temporal-Semantic Gate (ablation) | CaT-Diff · Time-MMD Energy / Public Health | MSE: Energy 0.147 -> 0.135; Public Health 18.487 -> 16.763 | Improved |
| use Cross-Key Ranking Attention (ablation) | CaT-Diff · Time-MMD Energy / Public Health | MSE: Energy 0.149 -> 0.135; Public Health 18.608 -> 16.763 | Improved |
| use Cross-Key Ranking Attention instead of standard cross-attention (ablation) | CaT-Diff · Time-MMD Energy / Public Health | MSE: Energy 0.138 -> 0.135; Public Health 17.110 -> 16.763 | Improved |
| use Cross-Key Ranking Attention instead of standard cross-attention (ablation) | CaT-Diff · Time-MMD overhead | GFLOPs: 1.38 -> 0.96 | Improved |
| add Mixture-of-Experts projection (ablation) | CaT-Diff · Time-MMD Energy / Public Health | MSE: Energy 0.156 -> 0.135; Public Health 19.396 -> 16.763 | Improved |
| add Missing Mechanism Estimator (ablation) | CaT-Diff · Time-MMD Energy / Public Health | MSE: Energy 0.141 -> 0.135; Public Health 18.359 -> 16.763 | Improved |
| add LLM text (ablation) | CaT-Diff · Time-MMD Energy / Public Health | MSE: Energy 0.170 -> 0.135; Public Health 21.680 -> 16.763 | Improved |
| use misaligned (shuffled) text instead of aligned text (ablation) | CaT-Diff · Time-MMD Energy / Public Health | MSE: Energy 0.170 -> 0.182; Public Health 21.680 -> 23.153 | Hurt |
| use semantically meaningless text instead of no text (ablation) | CaT-Diff · Time-MMD Energy / Public Health | MSE: Energy 0.170 -> 0.183; Public Health 21.680 -> 22.997 | Hurt |
| increase kappa_t beyond 48 (ablation) | CaT-Diff · Time-MMD Agriculture 50% mask | MSE | Hurt |
| increase kappa_en beyond 6 (ablation) | CaT-Diff · Time-MMD Agriculture 50% mask | MSE | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 5 on most shared results.
- TimeCMA — TimeCMA: Towards LLM-Empowered Multivariate Time Series Forecasting via Cross-Modality Alignment (AAAI 2025) · better on 5 of 5 shared results
- CSDI — CSDI: Conditional Score-based Diffusion Models for Probabilistic Time Series Imputation (NeurIPS 2021) · better on 4 of 4 shared results
- TimeLLM — Time-LLM: Time Series Forecasting by Reprogramming Large Language Models (ICLR 2024) · better on 1 of 2 shared results
- BRITS — BRITS: Bidirectional Recurrent Imputation for Time Series (NeurIPS 2018) · better on 1 of 2 shared results
- iTransformer — iTransformer: Inverted Transformers Are Effective for Time Series Forecasting (ICLR 2024) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Discretizing Continuous Time Series for Imputation with Masked Diffusion Training (NeurIPS 2026 · similarity 0.89)
- Diffusion Transformers for Imputation: Statistical Efficiency and Uncertainty Quantification (NeurIPS 2025 · similarity 0.89)
- Learning What Not to Impute: An Uncertainty-Aware Diffusion Framework for Meaningful Missingness (NeurIPS 2026 · similarity 0.89)
- Frequency-aware Generative Models for Multivariate Time Series Imputation (NeurIPS 2024 · similarity 0.88)
- ProCTI: Prototype-Refined Global Conditioning for Diffusion-Based Time Series Imputation (NeurIPS 2026 · similarity 0.88)
- DiffPuter: Empowering Diffusion Models for Missing Data Imputation (ICLR 2025 · similarity 0.87)
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.
