# Audio Reconstruction on LibriTTS, MTG-Jamendo, and Opencpop Average (test)

Source: [Audio Reconstruction on LibriTTS, MTG-Jamendo, and Opencpop Average (test)](<https://www.sota2.com/research/sota/audio-reconstruction-on-libritts-mtg-jamendo-and-opencpop-average-test>)

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

Data updated: 2026-02-26T05:50:05.065Z

Evaluation of audio reconstruction quality averaged across the test splits of LibriTTS, MTG-Jamendo, and Opencpop datasets, utilizing metrics such as MOS, PESQ, ViSQOL, V/UV F1, and Periodicity.

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 | Audio Reconstruction |
| Dataset (including any stated split) | LibriTTS, MTG-Jamendo, and Opencpop Average (test) |
| Metric | MOS |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 4 of 4 |
| 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

- [MOS](<https://www.sota2.com/research/sota/audio-reconstruction-on-libritts-mtg-jamendo-and-opencpop-average-test.md?metric=MOS>): Higher is better; unit: SCALAR.

- [PESQ](<https://www.sota2.com/research/sota/audio-reconstruction-on-libritts-mtg-jamendo-and-opencpop-average-test.md?metric=PESQ>): Higher is better; unit: SCALAR.

- [ViSQOL](<https://www.sota2.com/research/sota/audio-reconstruction-on-libritts-mtg-jamendo-and-opencpop-average-test.md?metric=ViSQOL>): Higher is better; unit: SCALAR.

- [V/UV F1](<https://www.sota2.com/research/sota/audio-reconstruction-on-libritts-mtg-jamendo-and-opencpop-average-test.md?metric=V%2FUV%20F1>): Higher is better; unit: PERCENT.

- [Periodicity Error](<https://www.sota2.com/research/sota/audio-reconstruction-on-libritts-mtg-jamendo-and-opencpop-average-test.md?metric=Periodicity%20Error>): Lower is better; unit: SCALAR.

## Recorded results

| Method | MOS | Result date | Report paper |
| --- | --- | --- | --- |
| Ground truth | 4 | Not recorded | 68fa75111e209131cea657ba |
| RFWave | 3.95 | Not recorded | 68fa75111e209131cea657ba |
| PriorGrad | 3.75 | Not recorded | 68fa75111e209131cea657ba |
| FreGrad | 2.99 | Not recorded | 68fa75111e209131cea657ba |

### Ground truth

MOS: 4. Result ID: 699371d53a6c31484915fd5d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75111e209131cea657ba.md>)

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

### RFWave

MOS: 3.95. Result ID: 699371d53a6c31484915fd5a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75111e209131cea657ba.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75111e209131cea657ba.md>)

[Implementation](<https://github.com/bfs18/rfwave>)

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

### PriorGrad

MOS: 3.75. Result ID: 699371d53a6c31484915fd5b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75111e209131cea657ba.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6bd31e209131cea305e0.md>)

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

### FreGrad

MOS: 2.99. Result ID: 699371d53a6c31484915fd5c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75111e209131cea657ba.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71d61e209131cea5344b.md>)

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

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