# Automatic Subtitling on EP Interviews (EPI) en-de (test)

Source: [Automatic Subtitling on EP Interviews (EPI) en-de (test)](<https://www.sota2.com/research/sota/automatic-subtitling-on-ep-interviews-epi-en-de-test>)

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

Data updated: 2026-02-26T06:04:51.939Z

Evaluation of Automatic Subtitling performance on the test split of the EP Interviews (EPI) English-to-German dataset, measured using SubER (cased) and SubSONAR.

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 | Automatic Subtitling |
| Dataset (including any stated split) | EP Interviews (EPI) en-de (test) |
| Metric | SubER (cased) |
| Unit | PERCENT |
| 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

- [SubER (cased)](<https://www.sota2.com/research/sota/automatic-subtitling-on-ep-interviews-epi-en-de-test.md?metric=SubER%20(cased)>): Higher is better; unit: PERCENT.

- [SubSONAR](<https://www.sota2.com/research/sota/automatic-subtitling-on-ep-interviews-epi-en-de-test.md?metric=SubSONAR>): Higher is better; unit: PERCENT.

## Recorded results

| Method | SubER (cased) | Result date | Report paper |
| --- | --- | --- | --- |
| Papi et al. (2023a) | 80.3 | Not recorded | 68fa73491e209131cea5b8d2 |
| Proposed (LEV) | 78.5 | Not recorded | 68fa73491e209131cea5b8d2 |
| Best Production | 78.1 | Not recorded | 68fa73491e209131cea5b8d2 |
| Proposed (LEV + Joint) | 77.6 | Not recorded | 68fa73491e209131cea5b8d2 |
| Proposed (DTW + Joint) | 77.3 | Not recorded | 68fa73491e209131cea5b8d2 |
| Proposed (SBAAM + Joint) | 77.3 | Not recorded | 68fa73491e209131cea5b8d2 |

### Papi et al. (2023a)

SubER (cased): 80.3. Result ID: 6992b68c3a6c31484910f073.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73491e209131cea5b8d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70271e209131cea496de.md>)

Recorded evaluation configuration: {"align":"LEV","joint":false}. This record does not assert independent reproduction.

### Proposed (LEV)

SubER (cased): 78.5. Result ID: 6992b68c3a6c31484910f075.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73491e209131cea5b8d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73491e209131cea5b8d2.md>)

[Implementation](<https://github.com/hlt-mt/FBK-fairseq/>)

Recorded evaluation configuration: {"align":"LEV","joint":false}. This record does not assert independent reproduction.

### Best Production

SubER (cased): 78.1. Result ID: 6992b68c3a6c31484910f074.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73491e209131cea5b8d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70271e209131cea496de.md>)

Recorded evaluation configuration: {"align":"na","joint":"na"}. This record does not assert independent reproduction.

### Proposed (LEV + Joint)

SubER (cased): 77.6. Result ID: 6992b68c3a6c31484910f076.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73491e209131cea5b8d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73491e209131cea5b8d2.md>)

[Implementation](<https://github.com/hlt-mt/FBK-fairseq/>)

Recorded evaluation configuration: {"align":"LEV","joint":true}. This record does not assert independent reproduction.

### Proposed (DTW + Joint)

SubER (cased): 77.3. Result ID: 6992b68c3a6c31484910f077.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73491e209131cea5b8d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73491e209131cea5b8d2.md>)

[Implementation](<https://github.com/hlt-mt/FBK-fairseq/>)

Recorded evaluation configuration: {"align":"DTW","joint":true}. This record does not assert independent reproduction.

### Proposed (SBAAM + Joint)

SubER (cased): 77.3. Result ID: 6992b68c3a6c31484910f078.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73491e209131cea5b8d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73491e209131cea5b8d2.md>)

[Implementation](<https://github.com/hlt-mt/FBK-fairseq/>)

Recorded evaluation configuration: {"align":"SBAAM","joint":true}. This record does not assert independent reproduction.

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