# Co-speech gesture generation on BEAT All Speakers 2

Source: [Co-speech gesture generation on BEAT All Speakers 2](<https://www.sota2.com/research/sota/co-speech-gesture-generation-on-beat-all-speakers-2>)

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

Data updated: 2026-03-30T07:57:14.513Z

Evaluation of Co-speech gesture generation models using quantitative metrics including FGD, BC, and Diversity on the BEAT All Speakers 2 dataset.

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 | Co-speech gesture generation |
| Dataset (including any stated split) | BEAT All Speakers 2 |
| Metric | BC |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 33 of 33 |
| 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

- [BC](<https://www.sota2.com/research/sota/co-speech-gesture-generation-on-beat-all-speakers-2.md?metric=BC>): Higher is better; unit: SCALAR.

- [FGD](<https://www.sota2.com/research/sota/co-speech-gesture-generation-on-beat-all-speakers-2.md?metric=FGD>): Higher is better; unit: SCALAR.

- [Diversity](<https://www.sota2.com/research/sota/co-speech-gesture-generation-on-beat-all-speakers-2.md?metric=Diversity>): Higher is better; unit: SCALAR.

- [FGD (B)](<https://www.sota2.com/research/sota/co-speech-gesture-generation-on-beat-all-speakers-2.md?metric=FGD%20(B)>): Higher is better; unit: SCALAR.

- [WER](<https://www.sota2.com/research/sota/co-speech-gesture-generation-on-beat-all-speakers-2.md?metric=WER>): Lower is better; unit: SCALAR.

- [NMOS](<https://www.sota2.com/research/sota/co-speech-gesture-generation-on-beat-all-speakers-2.md?metric=NMOS>): Higher is better; unit: SCALAR.

- [SS](<https://www.sota2.com/research/sota/co-speech-gesture-generation-on-beat-all-speakers-2.md?metric=SS>): Higher is better; unit: SCALAR.

## Recorded results

| Method | BC | Result date | Report paper |
| --- | --- | --- | --- |
| ViBES | 8.103 | Not recorded | 694238fa45614693628504f1 |
| SynTalker | 7.971 | Not recorded | 694238fa45614693628504f1 |
| LOM | 7.78 | Not recorded | 694238fa45614693628504f1 |
| EMAGE | 7.724 | Not recorded | 694238fa45614693628504f1 |
| Habibie et al. | 7.716 | Not recorded | 694238fa45614693628504f1 |
| DiffStyleGesture | 7.241 | Not recorded | 694238fa45614693628504f1 |
| TalkSHOW | 6.947 | Not recorded | 694238fa45614693628504f1 |
| CaMN | 6.769 | Not recorded | 694238fa45614693628504f1 |
| DisCo | 6.439 | Not recorded | 694238fa45614693628504f1 |
| Gelina - Flow | 0.824 | Not recorded | 68fa7c1e1e209131cea8c830 |
| HolisticSemGes | 0.78 | Not recorded | 69ca03a70636e5bf4afc948f |
| Gelina S2G | 0.768 | Not recorded | 68fa7c1e1e209131cea8c830 |
| EMAGE | 0.766 | Not recorded | 68fa7c1e1e209131cea8c830 |
| Gelina | 0.744 | Not recorded | 68fa7c1e1e209131cea8c830 |
| Gelina Clon. | 0.738 | Not recorded | 68fa7c1e1e209131cea8c830 |
| SemTalk | 0.727 | Not recorded | 69ca03a70636e5bf4afc948f |
| RAG | 0.7 | Not recorded | 68fa7c1e1e209131cea8c830 |
| EMAGE | 0.692 | Not recorded | 69ca03a70636e5bf4afc948f |
| Human | 0.684 | Not recorded | 68fa7c1e1e209131cea8c830 |
| Tokenizers | 0.667 | Not recorded | 68fa7c1e1e209131cea8c830 |
| SHOW | 0.652 | Not recorded | 69ca03a70636e5bf4afc948f |
| CAMN | 0.551 | Not recorded | 68fa7c1e1e209131cea8c830 |
| Intentional-Gesture | 0.534 | Not recorded | 68fa7ba41e209131cea89971 |
| GestureLSM | 0.525 | Not recorded | 68fa7ba41e209131cea89971 |
| GestureLSM | 0.525 | Not recorded | 69ca03a70636e5bf4afc948f |
| Ground Truth | 0.477 | Not recorded | 68fa7ba41e209131cea89971 |
| RAGGesture | 0.475 | Not recorded | 69ca03a70636e5bf4afc948f |
| RAG-Gesture | 0.471 | Not recorded | 68fa7ba41e209131cea89971 |
| Audio2Photoreal | 0.326 | Not recorded | 68fa7ba41e209131cea89971 |
| EMAGE | 0.284 | Not recorded | 68fa7ba41e209131cea89971 |
| CaMN | 0.2 | Not recorded | 68fa7ba41e209131cea89971 |

### ViBES

BC: 8.103. Result ID: 69983f685698665ec49bb920.

[Reporting paper](<https://www.sota2.com/research/paper/694238fa45614693628504f1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/694238fa45614693628504f1.md>)

Recorded evaluation configuration: {"training\_data":"BEATv2 training split only"}. This record does not assert independent reproduction.

### SynTalker

BC: 7.971. Result ID: 69983f685698665ec49bb91d.

[Reporting paper](<https://www.sota2.com/research/paper/694238fa45614693628504f1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75081e209131cea6565f.md>)

[Implementation](<https://robinwitch.github.io/SynTalker-Page/>)

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

### LOM

BC: 7.78. Result ID: 69983f685698665ec49bb91f.

[Reporting paper](<https://www.sota2.com/research/paper/694238fa45614693628504f1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76391e209131cea6bf5d.md>)

[Implementation](<https://www.sota2.com/languageofmotion.github.io>)

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

### EMAGE

BC: 7.724. Result ID: 69983f685698665ec49bb91e.

[Reporting paper](<https://www.sota2.com/research/paper/694238fa45614693628504f1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72c11e209131cea5867d.md>)

[Implementation](<https://pantomatrix.github.io/EMAGE/>)

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

### Habibie et al.

BC: 7.716. Result ID: 69983f685698665ec49bb91b.

[Reporting paper](<https://www.sota2.com/research/paper/694238fa45614693628504f1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69511e209131cea21863.md>)

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

### DiffStyleGesture

BC: 7.241. Result ID: 69983f685698665ec49bb91a.

[Reporting paper](<https://www.sota2.com/research/paper/694238fa45614693628504f1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f4d1e209131cea44eb3.md>)

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

### TalkSHOW

BC: 6.947. Result ID: 69983f685698665ec49bb91c.

[Reporting paper](<https://www.sota2.com/research/paper/694238fa45614693628504f1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6fcb1e209131cea4771a.md>)

[Implementation](<https://talkshow.is.tue.mpg.de/>)

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

### CaMN

BC: 6.769. Result ID: 69983f685698665ec49bb919.

[Reporting paper](<https://www.sota2.com/research/paper/694238fa45614693628504f1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d751e209131cea3a0d5.md>)

[Implementation](<https://pantomatrix.github.io/BEAT/>)

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

### DisCo

BC: 6.439. Result ID: 69983f685698665ec49bb918.

[Reporting paper](<https://www.sota2.com/research/paper/694238fa45614693628504f1.md>)

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

### Gelina - Flow

BC: 0.824. Result ID: 69ca197447f34ce924b20bbd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c1e1e209131cea8c830.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c1e1e209131cea8c830.md>)

Recorded evaluation configuration: {"Ablation":"Without flow-matching decoding"}. This record does not assert independent reproduction.

### HolisticSemGes

BC: 0.78. Result ID: 69ca2cc147f34ce924b29e2c.

[Reporting paper](<https://www.sota2.com/research/paper/69ca03a70636e5bf4afc948f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69ca03a70636e5bf4afc948f.md>)

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

### Gelina S2G

BC: 0.768. Result ID: 69ca197447f34ce924b20bbc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c1e1e209131cea8c830.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c1e1e209131cea8c830.md>)

Recorded evaluation configuration: {"Configuration":"Speech-to-Gesture","Input":"Ground-truth speech"}. This record does not assert independent reproduction.

### EMAGE

BC: 0.766. Result ID: 69ca197447f34ce924b20bb8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c1e1e209131cea8c830.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72c11e209131cea5867d.md>)

[Implementation](<https://pantomatrix.github.io/EMAGE/>)

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

### Gelina

BC: 0.744. Result ID: 69ca197447f34ce924b20bba.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c1e1e209131cea8c830.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c1e1e209131cea8c830.md>)

Recorded evaluation configuration: {"Configuration":"Base model"}. This record does not assert independent reproduction.

### Gelina Clon.

BC: 0.738. Result ID: 69ca197447f34ce924b20bbb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c1e1e209131cea8c830.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c1e1e209131cea8c830.md>)

Recorded evaluation configuration: {"Configuration":"Cloning","Conditioning":"Speech-gesture input"}. This record does not assert independent reproduction.

### SemTalk

BC: 0.727. Result ID: 69ca2cc147f34ce924b29e2b.

[Reporting paper](<https://www.sota2.com/research/paper/69ca03a70636e5bf4afc948f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77591e209131cea7292d.md>)

[Implementation](<https://xiangyue-zhang.github.io/SemTalk>)

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

### RAG

BC: 0.7. Result ID: 69ca197447f34ce924b20bb9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c1e1e209131cea8c830.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c11e209131cea751db.md>)

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

### EMAGE

BC: 0.692. Result ID: 69ca2cc147f34ce924b29e28.

[Reporting paper](<https://www.sota2.com/research/paper/69ca03a70636e5bf4afc948f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72c11e209131cea5867d.md>)

[Implementation](<https://pantomatrix.github.io/EMAGE/>)

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

### Human

BC: 0.684. Result ID: 69ca197447f34ce924b20bb5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c1e1e209131cea8c830.md>)

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

### Tokenizers

BC: 0.667. Result ID: 69ca197447f34ce924b20bb6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c1e1e209131cea8c830.md>)

Recorded evaluation configuration: {"Description":"Reconstruction from codecs without autoregressive modeling"}. This record does not assert independent reproduction.

### SHOW

BC: 0.652. Result ID: 69ca2cc147f34ce924b29e27.

[Reporting paper](<https://www.sota2.com/research/paper/69ca03a70636e5bf4afc948f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6fcb1e209131cea4771a.md>)

[Implementation](<https://talkshow.is.tue.mpg.de/>)

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

### CAMN

BC: 0.551. Result ID: 69ca197447f34ce924b20bb7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c1e1e209131cea8c830.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d751e209131cea3a0d5.md>)

[Implementation](<https://pantomatrix.github.io/BEAT/>)

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

### Intentional-Gesture

BC: 0.534. Result ID: 698b89157a967eddb76f62c2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7ba41e209131cea89971.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7ba41e209131cea89971.md>)

[Implementation](<https://andypinxinliu.github.io/Intentional-Gesture>)

Recorded evaluation configuration: {"speaker\_mode":"All Speakers"}. This record does not assert independent reproduction.

### GestureLSM

BC: 0.525. Result ID: 698b89157a967eddb76f62c1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7ba41e209131cea89971.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a641e209131cea82d92.md>)

[Implementation](<https://andypinxinliu.github.io/GestureLSM>)

Recorded evaluation configuration: {"speaker\_mode":"All Speakers"}. This record does not assert independent reproduction.

### GestureLSM

BC: 0.525. Result ID: 69ca2cc147f34ce924b29e2a.

[Reporting paper](<https://www.sota2.com/research/paper/69ca03a70636e5bf4afc948f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a641e209131cea82d92.md>)

[Implementation](<https://andypinxinliu.github.io/GestureLSM>)

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

### Ground Truth

BC: 0.477. Result ID: 698b89157a967eddb76f62bc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7ba41e209131cea89971.md>)

Recorded evaluation configuration: {"speaker\_mode":"All Speakers"}. This record does not assert independent reproduction.

### RAGGesture

BC: 0.475. Result ID: 69ca2cc147f34ce924b29e29.

[Reporting paper](<https://www.sota2.com/research/paper/69ca03a70636e5bf4afc948f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c11e209131cea751db.md>)

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

### RAG-Gesture

BC: 0.471. Result ID: 698b89157a967eddb76f62c0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7ba41e209131cea89971.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c11e209131cea751db.md>)

Recorded evaluation configuration: {"speaker\_mode":"All Speakers"}. This record does not assert independent reproduction.

### Audio2Photoreal

BC: 0.326. Result ID: 698b89157a967eddb76f62bf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7ba41e209131cea89971.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52ca0.md>)

Recorded evaluation configuration: {"speaker\_mode":"All Speakers"}. This record does not assert independent reproduction.

### EMAGE

BC: 0.284. Result ID: 698b89157a967eddb76f62be.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7ba41e209131cea89971.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72c11e209131cea5867d.md>)

[Implementation](<https://pantomatrix.github.io/EMAGE/>)

Recorded evaluation configuration: {"speaker\_mode":"All Speakers"}. This record does not assert independent reproduction.

### CaMN

BC: 0.2. Result ID: 698b89157a967eddb76f62bd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7ba41e209131cea89971.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d751e209131cea3a0d5.md>)

[Implementation](<https://pantomatrix.github.io/BEAT/>)

Recorded evaluation configuration: {"speaker\_mode":"All Speakers"}. This record does not assert independent reproduction.

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