Research map / Probabilistic and scientific modeling / Molecules and proteins
Physics-Informed Audio-Geometry-Grid Representation Learning for Universal Sound Source Localization
Min-Sang Baek, Gyeong-Su Kim, Donghyun Kim, Joon-Hyuk Chang
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 LNuDFT-based GCC-PHAT and rMPE in place of DFT with aMPE (ablation) | AuGeonet · Eigenmike | MAE (deg): 111.21 -> 11.24 | Improved |
| add learnable non-uniform DFT in place of standard DFT (ablation) | AuGeonet · Eigenmike | MAE (deg): 16.53 -> 11.24 | Improved |
| add learnable non-uniform DFT in place of standard DFT (ablation) | AuGeonet · Dynamic-S | MAE (deg): 10.26 -> 10.32 | No clear effect |
| use logit-based init instead of uniform init for LNuDFT (ablation) | AuGeonet · Dynamic-U | MAE (deg): 23.03 -> 14.12 | Improved |
| make NuDFT bins learnable rather than frozen (ablation) | AuGeonet · Eigenmike | MAE (deg): 17.34 -> 11.24 | Improved |
| make NuDFT bins learnable rather than frozen (ablation) | AuGeonet · Dynamic-U | MAE (deg): 11.83 -> 14.12 | Hurt |
| use AGG-RL grid representation instead of fixed grid (ablation) | AuGeonet · Eigenmike | MAE (deg): 13.58 -> 11.24 | Improved |
| use AGG-RL grid representation instead of fixed grid (ablation) | AuGeonet · Dynamic-S | MAE (deg): 9.57 -> 10.32 | Hurt |
| use FM-based rMPE instead of PM-based rMPE (ablation) | AuGeonet · Dynamic-U | MAE (deg): 12.46 -> 14.12 | Hurt |
| apply AGG-RL to Neural-SRP | Neural-SRP · Eigenmike | MAE (deg): 52.75 -> 13.25 | Improved |
| apply AGG-RL to Unet | Unet · Eigenmike | MAE (deg): 14.89 -> 16.86 | Hurt |
| use sinusoidal grid encoding instead of raw Cartesian input to Gridnet (ablation) | AuGeonet · Dynamic-U | MAE (deg): 23.10 -> 14.12 | Improved |
Closest papers by meaning
- Representing Sounds as Neural Amplitude Fields: A Benchmark of Coordinate-MLPs and a Fourier Kolmogorov-Arnold Framework (AAAI 2025 · similarity 0.87)
- OWL : Geometry-Aware Spatial Reasoning for Audio Large Language Models (ICLR 2026 · similarity 0.87)
- JAEGER: Joint 3D Audio-Visual Grounding and Reasoning in Simulated Physical Environments (ICML 2026 · similarity 0.86)
- MRSAudio: A Large-Scale Multimodal Recorded Spatial Audio Dataset with Refined Annotations (NeurIPS 2025 · similarity 0.85)
- Improving Sound Source Localization with Joint Slot Attention on Image and Audio (CVPR 2025 · similarity 0.85)
- BAT: Learning to Reason about Spatial Sounds with Large Language Models (ICML 2024 · similarity 0.85)
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.
