# Group Dance Generation on AIOZ-GDance long-duration (test)

Source: [Group Dance Generation on AIOZ-GDance long-duration (test)](<https://www.sota2.com/research/sota/group-dance-generation-on-aioz-gdance-long-duration-test>)

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

Data updated: 2026-03-25T12:46:16.199Z

Evaluation of models generating group dance sequences using the long-duration test split of the AIOZ-GDance dataset, assessed using metrics including GMR, GMC, TIF, FID, Div, and PFC.

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 | Group Dance Generation |
| Dataset (including any stated split) | AIOZ-GDance long-duration (test) |
| Metric | GMR |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 6 of 6 |
| 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

- [GMR](<https://www.sota2.com/research/sota/group-dance-generation-on-aioz-gdance-long-duration-test.md?metric=GMR>): Higher is better; unit: SCALAR.

- [GMC](<https://www.sota2.com/research/sota/group-dance-generation-on-aioz-gdance-long-duration-test.md?metric=GMC>): Higher is better; unit: SCALAR.

- [TIF](<https://www.sota2.com/research/sota/group-dance-generation-on-aioz-gdance-long-duration-test.md?metric=TIF>): Higher is better; unit: PERCENT.

- [FID](<https://www.sota2.com/research/sota/group-dance-generation-on-aioz-gdance-long-duration-test.md?metric=FID>): Lower is better; unit: SCALAR.

- [Div](<https://www.sota2.com/research/sota/group-dance-generation-on-aioz-gdance-long-duration-test.md?metric=Div>): Higher is better; unit: SCALAR.

- [PFC](<https://www.sota2.com/research/sota/group-dance-generation-on-aioz-gdance-long-duration-test.md?metric=PFC>): Higher is better; unit: SCALAR.

## Recorded results

| Method | GMR | Result date | Report paper |
| --- | --- | --- | --- |
| EDGE | 67.24 | Not recorded | 69c36c190636e5bf4afc6517 |
| GCD | 40.68 | Not recorded | 69c36c190636e5bf4afc6517 |
| CoDancers | 35.2 | Not recorded | 69c36c190636e5bf4afc6517 |
| ST-GDance | 34.25 | Not recorded | 69c36c190636e5bf4afc6517 |
| TCDiff++ | 14.67 | Not recorded | 69c36c190636e5bf4afc6517 |
| ST-GDance++ | 14.32 | Not recorded | 69c36c190636e5bf4afc6517 |

### EDGE

GMR: 67.24. Result ID: 69c3d88947f34ce924ad1ca5.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c190636e5bf4afc6517.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e0d1e209131cea3db38.md>)

Recorded evaluation configuration: {"FLOPs (G)":"68.21","Params (M)":"55.27","Train (min/epoch)":"2:56","Inf (s)":"104.6"}. This record does not assert independent reproduction.

### GCD

GMR: 40.68. Result ID: 69c3d88947f34ce924ad1ca6.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c190636e5bf4afc6517.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa710f1e209131cea4ef4e.md>)

[Implementation](<https://aioz-ai.github.io/GCD/>)

Recorded evaluation configuration: {"FLOPs (G)":"92.93","Params (M)":"64.42","Train (min/epoch)":"4:12","Inf (s)":"92.4"}. This record does not assert independent reproduction.

### CoDancers

GMR: 35.2. Result ID: 69c3d88947f34ce924ad1ca7.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c190636e5bf4afc6517.md>)

Recorded evaluation configuration: {"FLOPs (G)":"153.70","Params (M)":"59.32","Train (min/epoch)":"5:23","Inf (s)":"72.3"}. This record does not assert independent reproduction.

### ST-GDance

GMR: 34.25. Result ID: 69c3d88947f34ce924ad1ca9.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c190636e5bf4afc6517.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a451e209131cea82167.md>)

Recorded evaluation configuration: {"FLOPs (G)":"15.92","Params (M)":"51.42","Train (min/epoch)":"1:31","Inf (s)":"30.5"}. This record does not assert independent reproduction.

### TCDiff++

GMR: 14.67. Result ID: 69c3d88947f34ce924ad1ca8.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c190636e5bf4afc6517.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b2a3.md>)

Recorded evaluation configuration: {"FLOPs (G)":"168.72","Params (M)":"72.64","Train (min/epoch)":"5:43","Inf (s)":"124.4"}. This record does not assert independent reproduction.

### ST-GDance++

GMR: 14.32. Result ID: 69c3d88947f34ce924ad1caa.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c190636e5bf4afc6517.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c36c190636e5bf4afc6517.md>)

Recorded evaluation configuration: {"FLOPs (G)":"29.32","Params (M)":"52.83","Train (min/epoch)":"1:45","Inf (s)":"32.5"}. This record does not assert independent reproduction.

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