# Visual Text-to-Speech on Chem

Source: [Visual Text-to-Speech on Chem](<https://www.sota2.com/research/sota/visual-text-to-speech-on-chem>)

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

Data updated: 2026-03-11T06:42:25.208Z

Evaluation of Visual Text-to-Speech synthesis quality on the Chem dataset, measured using metrics such as Word Error Rate (WER), Speaker Similarity (Spk. Sim.), UTMOS, Mean Cepstral Distortion (MCD variants), and LSE scores.

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 | Visual Text-to-Speech |
| Dataset (including any stated split) | Chem |
| Metric | WER |
| Unit | PERCENT |
| Direction | Lower is better |
| Results loaded | 10 of 10 |
| 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

- [WER](<https://www.sota2.com/research/sota/visual-text-to-speech-on-chem.md?metric=WER>): Lower is better; unit: PERCENT.

- [Speaker Similarity (Spk. Sim.)](<https://www.sota2.com/research/sota/visual-text-to-speech-on-chem.md?metric=Speaker%20Similarity%20(Spk.%20Sim.)>): Higher is better; unit: PERCENT.

- [UTMOS Score](<https://www.sota2.com/research/sota/visual-text-to-speech-on-chem.md?metric=UTMOS%20Score>): Higher is better; unit: SCALAR.

- [Mean Cepstral Distortion (MCD)](<https://www.sota2.com/research/sota/visual-text-to-speech-on-chem.md?metric=Mean%20Cepstral%20Distortion%20(MCD)>): Lower is better; unit: SCALAR.

- [MCD (D variant)](<https://www.sota2.com/research/sota/visual-text-to-speech-on-chem.md?metric=MCD%20(D%20variant)>): Lower is better; unit: SCALAR.

- [MCD (DS variant)](<https://www.sota2.com/research/sota/visual-text-to-speech-on-chem.md?metric=MCD%20(DS%20variant)>): Lower is better; unit: SCALAR.

- [LSE Score (C variant)](<https://www.sota2.com/research/sota/visual-text-to-speech-on-chem.md?metric=LSE%20Score%20(C%20variant)>): Lower is better; unit: SCALAR.

- [LSE Score (D variant)](<https://www.sota2.com/research/sota/visual-text-to-speech-on-chem.md?metric=LSE%20Score%20(D%20variant)>): Lower is better; unit: SCALAR.

## Recorded results

| Method | WER | Result date | Report paper |
| --- | --- | --- | --- |
| GT | 3.5 | Not recorded | 68fa7bc71e209131cea8a6b9 |
| GT-vocoder | 3.5 | Not recorded | 68fa7bc71e209131cea8a6b9 |
| E2-TTS | 8.7 | Not recorded | 68fa7bc71e209131cea8a6b9 |
| VSSFlow-L | 9.2 | Not recorded | 68fa7bc71e209131cea8a6b9 |
| VSSFlow-M | 9.4 | Not recorded | 68fa7bc71e209131cea8a6b9 |
| VSSFlow-S | 12.6 | Not recorded | 68fa7bc71e209131cea8a6b9 |
| StyleDubber | 16.3 | Not recorded | 68fa7bc71e209131cea8a6b9 |
| EmoDubber | 16.7 | Not recorded | 68fa7bc71e209131cea8a6b9 |
| HPMDubbing | 22.3 | Not recorded | 68fa7bc71e209131cea8a6b9 |
| DSU | 37.3 | Not recorded | 68fa7bc71e209131cea8a6b9 |

### GT

WER: 3.5. Result ID: 69b10df26dbb7075b95ac1fb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc71e209131cea8a6b9.md>)

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

### GT-vocoder

WER: 3.5. Result ID: 69b10df26dbb7075b95ac1fc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc71e209131cea8a6b9.md>)

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

### E2-TTS

WER: 8.7. Result ID: 69b10df26dbb7075b95ac201.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc71e209131cea8a6b9.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74cc1e209131cea63eaf.md>)

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

### VSSFlow-L

WER: 9.2. Result ID: 69b10df26dbb7075b95ac204.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc71e209131cea8a6b9.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bc71e209131cea8a6b9.md>)

Recorded evaluation configuration: {"Model Size":"Large"}. This record does not assert independent reproduction.

### VSSFlow-M

WER: 9.4. Result ID: 69b10df26dbb7075b95ac203.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc71e209131cea8a6b9.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bc71e209131cea8a6b9.md>)

Recorded evaluation configuration: {"Model Size":"Medium"}. This record does not assert independent reproduction.

### VSSFlow-S

WER: 12.6. Result ID: 69b10df26dbb7075b95ac202.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc71e209131cea8a6b9.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bc71e209131cea8a6b9.md>)

Recorded evaluation configuration: {"Model Size":"Small"}. This record does not assert independent reproduction.

### StyleDubber

WER: 16.3. Result ID: 69b10df26dbb7075b95ac1ff.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc71e209131cea8a6b9.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73f11e209131cea5f044.md>)

[Implementation](<https://github.com/GalaxyCong/StyleDubber>)

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

### EmoDubber

WER: 16.7. Result ID: 69b10df26dbb7075b95ac200.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc71e209131cea8a6b9.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78331e209131cea772b4.md>)

[Implementation](<https://galaxycong.github.io/EmoDub>)

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

### HPMDubbing

WER: 22.3. Result ID: 69b10df26dbb7075b95ac1fe.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc71e209131cea8a6b9.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ef51e209131cea432e1.md>)

[Implementation](<https://github.com/GalaxyCong/HPMDubbing>)

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

### DSU

WER: 37.3. Result ID: 69b10df26dbb7075b95ac1fd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc71e209131cea8a6b9.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6feb1e209131cea4828b.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.