Phoneme Decoding on Willett 2023 (test)
14.3PERTransformer Seq2Seq + MFCC + BART
Evaluation results
| Method | Links | |
|---|---|---|
| Transformer Seq2Seq + MFCC + BARTBase architecture=Transformer Seq2Seq, Auxiliary acoustic features (MFCC)=true, Word-level multitask objective (BART)=true2026.03 | 14.3 | |
| Transformer Seq2Seq + MFCC + BART + rescoringBase architecture=Transformer Seq2Seq, Auxiliary acoustic features (MFCC)=true, Word-level multitask objective (BART)=true, Post-processing/Rescoring=true2026.03 | 14.3 | |
| Transformer Seq2Seq + BARTBase architecture=Transformer Seq2Seq, Word-level multitask objective (BART)=true2026.03 | 14.5 | |
| Transformer Seq2Seq + MFCCBase architecture=Transformer Seq2Seq, Auxiliary acoustic features (MFCC)=true2026.03 | 14.6 | |
| Transformer Seq2SeqBase architecture=Transformer Seq2Seq2026.03 | 14.8 | |
| DCoND-LIFT2026.03 | 15.34 | |
| RNN–CTC + MFCCBase architecture=RNN-CTC, Auxiliary acoustic features (MFCC)=true2026.03 | 17 | |
| RNN–CTC + MFCC + BARTBase architecture=RNN-CTC, Auxiliary acoustic features (MFCC)=true, Word-level multitask objective (BART)=true2026.03 | 17 | |
| Transformer Seq2Seq + BART + Linear DTBase architecture=Transformer Seq2Seq, Word-level multitask objective (BART)=true, Day transform protocol=Linear DT2026.03 | 17.2 | |
| RNN–CTCBase architecture=RNN-CTC2026.03 | 17.4 | |
| Transformer Seq2Seq + BART + No DTBase architecture=Transformer Seq2Seq, Word-level multitask objective (BART)=true, Day transform protocol=No DT2026.03 | 18.1 | |
| RNN–CTC + LMBase architecture=RNN-CTC, Post-processing/Rescoring=Language Model (LM)2026.03 | 19.7 | |
| RNN–CTC + BARTBase architecture=RNN-CTC, Word-level multitask objective (BART)=true2026.03 | 21 |