Spoken Language Glossification on PHOENIX14T (test)
25.7BLEU-4S3LG
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| S3LGInference Strategy=Beam search, Search width=3, Length penalty=12024.06 | 25.7 | 59.62 | 60.67 | 43.69 | 32.91 | |
| Zhu et al.2024.06 | 24.89 | — | — | — | — | |
| Walsh et al.2024.06 | 23.19 | 56.55 | 58.74 | 40.86 | 30.24 | |
| Egea Gómez et al.year=20222024.06 | 20.57 | — | — | — | — | |
| BaselineArchitecture=Vanilla Transformer2024.06 | 20.22 | 55.28 | 57.36 | 38.9 | 27.8 | |
| Saunders et al.year=20222024.06 | 20.08 | 56.63 | — | — | — | |
| Saunders et al.year=20202024.06 | 19.1 | 54.55 | 55.18 | 37.1 | 26.24 | |
| Li et al.2024.06 | 18.89 | 49.91 | — | — | 25.51 | |
| Zhang and Duh2024.06 | 16.43 | — | — | — | — | |
| Stoll et al.2024.06 | 15.26 | 48.1 | 50.67 | 32.25 | 21.54 | |
| Egea Gómez et al.year=20212024.06 | 13.13 | — | — | — | — | |
| Amin et al.2024.06 | 10.42 | 42.96 | 43.9 | 26.33 | 16.16 |