Research map / Image generation / Image restoration
TS-DDAE: A novel Temporal-Spectral Denoising Diffusion AutoEncoder for Wireless Signal Recognition Model Pre-training
Yaoqi Liu, Jin Wang, Hui Wang, Chuan Shi
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 the temporal encoder (ablation) | TS-DDAE · RML2016.10A | overall accuracy: 48.81 -> 63.61 | Improved |
| add the temporal encoder (ablation) | TS-DDAE · RML2016.10B | overall accuracy: 50.94 -> 65.50 | Improved |
| add the spectral encoder (ablation) | TS-DDAE · TechRec | overall accuracy: 37.48 -> 89.62 | Improved |
| add the spectral encoder (ablation) | TS-DDAE · RML2016.10A | overall accuracy: 62.97 -> 63.61 | Improved |
| add the interactive process between temporal and spectral encoders (ablation) | TS-DDAE · TechRec | overall accuracy: 87.87 -> 89.62 | Improved |
| add the interactive process between temporal and spectral encoders (ablation) | TS-DDAE · RML2016.10B | overall accuracy: 65.09 -> 65.50 | Improved |
| use two noise levels instead of a single noise (ablation) | TS-DDAE · RML2016.10A | overall accuracy: 63.01 -> 63.61 | Improved |
| use two noise levels instead of a single noise (ablation) | TS-DDAE · TechRec | overall accuracy: 89.27 -> 89.62 | Improved |
| vary noise balance lambda from 0.2 to 1.0 (ablation) | TS-DDAE · RML2016.10A | accuracy: 62.22 -> 63.23 | Mixed |
| fine-tune the whole model instead of linear probe | TS-DDAE · TechRec | average accuracy: 86.00 -> 89.62 | Improved |
| fine-tune with full training data instead of 100-shot | TS-DDAE · RML2016.10A | average accuracy: 55.47 -> 63.61 | Improved |
| use TS-DDAE pre-training instead of IQFormer | TS-DDAE · RML2018 | average accuracy: 40.22 -> 64.15 | Improved |
Closest papers by meaning
- Deconstructing Denoising Diffusion Models for Self-Supervised Learning (ICLR 2025 · similarity 0.82)
- RestoreGrad: Signal Restoration Using Conditional Denoising Diffusion Models with Jointly Learned Prior (ICML 2025 · similarity 0.80)
- Content-aware Balanced Spectrum Encoding in Masked Modeling for Time Series Classification (AAAI 2025 · similarity 0.80)
- Diffusion-TS: Interpretable Diffusion for General Time Series Generation (ICLR 2024 · similarity 0.80)
- Denoising Task Routing for Diffusion Models (ICLR 2024 · similarity 0.80)
- Robust Automatic Modulation Classification with Fuzzy Regularization (ICML 2025 · similarity 0.80)
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.
