Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
End-to-End Probabilistic Inference for Nonstationary Audio Analysis
William Wilkinson, Michael Andersen, Joshua D. Reiss, Dan Stowell, Arno Solin
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 |
|---|---|---|---|
| Use 20 EP iterations instead of a single sweep (ablation) | GTF-NMF with EP · Simulated data | RMSE: 0.044 -> 0.003 | Improved |
| Use EP inference instead of extended Kalman filtering | GTF-NMF · Simulated data | RMSE: 0.128 -> 0.003 | Improved |
| Use 20 EP iterations instead of a single sweep (ablation) | GTF-NMF with EP · Missing data imputation | SNR: 7.494 -> 8.087 | Improved |
| Use 20 EP iterations instead of a single sweep (ablation) | GTF-NMF with EP · Missing data imputation | RMSE: 0.590 -> 0.551 | Improved |
| Use EP inference instead of extended Kalman filtering | GTF-NMF · Missing data imputation | SNR: 3.735 -> 8.087 | Improved |
| Use EP inference instead of extended Kalman filtering | GTF-NMF · Missing data imputation | RMSE: 0.743 -> 0.551 | Improved |
| Use EP20 instead of matching pursuit | Missing data imputation · Musical instrument recordings | SNR: 5.232 -> 8.087 | Improved |
| Use EP20 instead of matching pursuit | Missing data imputation · Musical instrument recordings | RMSE: 0.761 -> 0.551 | Improved |
| Iterate EP inference with the infinite-horizon GP approximation (ablation) | GTF-NMF with IHGP · Simulated data | RMSE: 0.042 -> 0.029 | Improved |
| Iterate EP inference with the infinite-horizon GP approximation (ablation) | GTF-NMF with IHGP · Missing data imputation | SNR: 4.520 -> 4.591 | Improved |
| Iterate EP inference with the infinite-horizon GP approximation (ablation) | GTF-NMF with IHGP · Missing data imputation | RMSE: 0.720 -> 0.716 | Improved |
| Use 20 EKF iterations instead of a single sweep (ablation) | GTF-NMF with EKF · Simulated data | RMSE: 0.124 -> 0.128 | Hurt |
Closest papers by meaning
- Non-Stationary Spectral Kernels (NeurIPS 2017 · similarity 0.85)
- Latent Gaussian Activity Propagation: Using Smoothness and Structure to Separate and Localize Sounds in Large Noisy Environments (NeurIPS 2018 · similarity 0.85)
- Revisiting Nonstationary Kernel Design for Multi-Output Gaussian Processes (ICLR 2026 · similarity 0.84)
- Scalable Gaussian Process Separation for Kernels with a Non-Stationary Phase (ICML 2020 · similarity 0.84)
- Nonstationary Sparse Spectral Permanental Process (NeurIPS 2024 · similarity 0.83)
- Representing Sounds as Neural Amplitude Fields: A Benchmark of Coordinate-MLPs and a Fourier Kolmogorov-Arnold Framework (AAAI 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.
