Research map / Video / Video generation
Audeo: Audio Generation for a Silent Performance Video
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Refine Video2Roll predictions with Roll2Midi Net | Audeo · Midi Evaluation Set | F1-score (%): 56.4 -> 61.5 | Helped |
| Use threshold 0.5 instead of 0.4 for Roll2Midi Net (ablation) | Roll2Midi Net · Midi Evaluation Set | F1-score (%): 61.5 -> 60.4 | Hurt |
| Use threshold 0.5 instead of 0.4 for Roll2Midi Net (ablation) | Roll2Midi Net · Midi Evaluation Set | precision (%): 60.0 -> 65.1 | Helped |
| Use threshold 0.5 instead of 0.4 for Roll2Midi Net (ablation) | Roll2Midi Net · Midi Evaluation Set | recall (%): 77.0 -> 69.9 | Hurt |
| Use threshold 0.5 instead of 0.4 for Roll2Midi Net (ablation) | Roll2Midi Net · Midi Evaluation Set | accuracy (%): 50.6 -> 50.8 | Helped |
| Synthesize refined Midi with FluidSynth rather than PerfNet (ablation) | Audeo · Audio Evaluation Set (Total) | SoundHound detection rate (%): 73.0 -> 73.9 | Helped |
| Use refined Midi rather than raw Roll as the synthesis input (ablation) | Audeo with FluidSynth · Audio Evaluation Set (Total) | SoundHound detection rate (%): 62.6 -> 73.9 | Helped |
| Use refined Midi rather than ResNet output as the synthesis input | Audeo with FluidSynth · Audio Evaluation Set (Total) | SoundHound detection rate (%): 55.9 -> 73.9 | Helped |
| Synthesize refined Midi with FluidSynth rather than PerfNet (ablation) | Audeo · Bach WTC B2 & Other | SoundHound detection rate (%): 71.6 -> 72.4 | Helped |
| Synthesize refined Midi with FluidSynth rather than PerfNet (ablation) | Audeo · Bach WTC B1 Variants | SoundHound detection rate (%): 80.6 -> 85.6 | Helped |
Closest papers by meaning
- Rubato: Transcribing Piano Music with Timestamps (NeurIPS 2026 · similarity 0.87)
- Aria-MIDI: A Dataset of Piano MIDI Files for Symbolic Music Modeling (ICLR 2025 · similarity 0.87)
- V2Meow: Meowing to the Visual Beat via Video-to-Music Generation (AAAI 2024 · similarity 0.86)
- VidMuse: A Simple Video-to-Music Generation Framework with Long-Short-Term Modeling (CVPR 2025 · similarity 0.85)
- MIDI-GPT: A Controllable Generative Model for Computer-Assisted Multitrack Music Composition (AAAI 2025 · similarity 0.84)
- PianoMotion10M: Dataset and Benchmark for Hand Motion Generation in Piano Performance (ICLR 2025 · similarity 0.84)
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.