Model configurations and descriptions.
| Model | Description |
|---|---|
| Baseline models | |
| Clean model baseline | Trained on clean transcriptions |
| Noisy model baseline | Trained on noisy transcriptions |
| TQL-aware models | |
| w/ UTT TQL | With utterance-level TQL |
| w/ CHAR TQL | With character-level TQL |
| w/ UTT TQL + CHAR TQL | With utterance and character TQL |
| Position-aware models | |
| w/ APE | With absolute positional encoding |
| w/ P’UTT-TQL | With APE and utterance-level TQL |
| w/ P’CHAR-TQL | With APE and character-level TQL |
| w/ P’UTT-TQL + CHAR-TQL | With APE and combined TQL |
| Model | Description |
|---|---|
| Clean model baseline | Trained on clean transcriptions |
| Noisy model baseline | Trained on noisy transcriptions |
| w/ | With utterance-level |
| w/ | With character-level |
| w/ | With utterance and character |
| w/ | With absolute positional encoding |
| w/ P’UTT-TQL | With |
| w/ P’CHAR-TQL | With |
| w/ P’UTT-TQL + CHAR-TQL | With |
Note(s): TQL- and position-aware models are trained on noisy transcriptions
Sharing content requires targeting cookies to be enabled. Please update your cookie preferences to use this feature.