# Expressive Performance Rendering on Piano (test)

Source: [Expressive Performance Rendering on Piano (test)](<https://www.sota2.com/research/sota/expressive-performance-rendering-on-piano-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-06-11T11:18:05.946Z

Evaluation of models performing Expressive Performance Rendering on the test split of the Piano dataset, measured using FAD, KAD, Pitch DTW, Alignment Precision, and Alignment Recall.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Expressive Performance Rendering |
| Dataset (including any stated split) | Piano (test) |
| Metric | Pitch DTW |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 3 of 3 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Pitch DTW](<https://www.sota2.com/research/sota/expressive-performance-rendering-on-piano-test.md?metric=Pitch%20DTW>): Lower is better; unit: SCALAR.

- [FAD](<https://www.sota2.com/research/sota/expressive-performance-rendering-on-piano-test.md?metric=FAD>): Lower is better; unit: SCALAR.

- [KAD](<https://www.sota2.com/research/sota/expressive-performance-rendering-on-piano-test.md?metric=KAD>): Lower is better; unit: SCALAR.

- [Alignment Precision](<https://www.sota2.com/research/sota/expressive-performance-rendering-on-piano-test.md?metric=Alignment%20Precision>): Higher is better; unit: PERCENT.

- [Alignment Recall](<https://www.sota2.com/research/sota/expressive-performance-rendering-on-piano-test.md?metric=Alignment%20Recall>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Pitch DTW | Result date | Report paper |
| --- | --- | --- | --- |
| CFG Bridge | 0.856 | Not recorded | 6a2a4151d33dfb23e7839569 |
| Human | 0.883 | Not recorded | 6a2a4151d33dfb23e7839569 |
| PianoKontext | 0.888 | Not recorded | 6a2a4151d33dfb23e7839569 |

### CFG Bridge

Pitch DTW: 0.856. Result ID: 6a2a9950fbc373592c28cf55.

[Reporting paper](<https://www.sota2.com/research/paper/6a2a4151d33dfb23e7839569.md>)

Recorded evaluation configuration: {"ODE solver":"Heun","Steps":"64","Guidance scale (inversion)":"1.0","Guidance scale (denoising)":"2.0","Context":"First 11s"}. This record does not assert independent reproduction.

### Human

Pitch DTW: 0.883. Result ID: 6a2a9950fbc373592c28cf57.

[Reporting paper](<https://www.sota2.com/research/paper/6a2a4151d33dfb23e7839569.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### PianoKontext

Pitch DTW: 0.888. Result ID: 6a2a9950fbc373592c28cf56.

[Reporting paper](<https://www.sota2.com/research/paper/6a2a4151d33dfb23e7839569.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a2a4151d33dfb23e7839569.md>)

[Implementation](<https://github.com/realfolkcode/pianokontext>)

Recorded evaluation configuration: {"ODE solver":"Heun","Steps":"64"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.