Research map / 3D vision / 3D generation
Be Everywhere - Hear Everything (BEE): Audio Scene Reconstruction by Sparse Audio-Visual Samples
Mingfei Chen, Kun Su, Eli Shlizerman
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 joint audio-visual rendering from sparse receivers instead of best prior learning method (APNet) | BEE · Matterport3D unseen scenes | STFT error: 0.515 -> 0.438 | Improved |
| add Audio-Visual Spatial Encoder (ablation) | BEE · Matterport3D unseen scenes | ENV: 0.512 -> 0.458 | Improved |
| add Spatial Enhanced Visual Representation (visual attention) (ablation) | BEE · Matterport3D unseen scenes | STFT error: 0.454 -> 0.438 | Improved |
| add cross-attention and averaged integrated rendering over receivers (ablation) | BEE · Matterport3D unseen scenes | STFT error: 0.509 -> 0.438 | Improved |
| add transposed-conv audio-visual branch (ablation) | BEE · Matterport3D unseen scenes | DPAM: 0.424 -> 0.348 | Improved |
| add MLP magnitude branch (ablation) | BEE · Matterport3D unseen scenes | STFT error: 0.626 -> 0.438 | Improved |
| use BEE vs non-learning baselines | BEE · Replica and Matterport3D | STFT error: 3.456 -> 0.438 (Mean vs BEE, Matterport3D unseen) | Improved |
| use BEE for human preference | BEE · human subject study | preference votes (%): 29.6% / 26.4% -> 44% | Improved |
Closest papers by meaning
- Hearing Anything Anywhere (CVPR 2024 · similarity 0.89)
- Audible World Models: Spatially Aware Sound Generation for 3D Worlds (NeurIPS 2026 · similarity 0.88)
- SoundVista: Novel-View Ambient Sound Synthesis via Visual-Acoustic Binding (CVPR 2025 · similarity 0.88)
- AV-Cloud: Spatial Audio Rendering Through Audio-Visual Cloud Splatting (NeurIPS 2024 · similarity 0.87)
- Resounding Acoustic Fields with Reciprocity (NeurIPS 2025 · similarity 0.87)
- NeRAF: 3D Scene Infused Neural Radiance and Acoustic Fields (ICLR 2025 · similarity 0.86)
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.
