Automatic Speech Recognition on MLS
5.26WERCrisperWhisper
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| CrisperWhisper2026.04 | 5.26 | — | — | — | — | |
| Mixed Training + Length Aug + Reduced TFLength Augmentation=true, Reduced Teacher Forcing=true, Evaluation Setting=zero-shot2026.04 | 5.63 | — | — | — | — | |
| Mixed Training + Length AugLength Augmentation=true, Evaluation Setting=zero-shot2026.04 | 5.68 | — | — | — | — | |
| Mixed Training + Length Aug + Timestamp RegLength Augmentation=true, Timestamp Regularization=true, Evaluation Setting=zero-shot2026.04 | 5.69 | — | — | — | — | |
| Mixed TrainingEvaluation Setting=zero-shot2026.04 | 5.81 | — | — | — | — | |
| ASR-only BaselineEvaluation Setting=zero-shot2026.04 | 6.84 | — | — | — | — | |
| Qwen-Audio2026.04 | 7.54 | — | — | — | — | |
| Gemini-1.5 Pro2024.07 | — | 4.6 | — | — | — | |
| Google USM2024.07 | — | 7 | 4.2 | 7.3 | 6.8 | |
| Seed-ASR (ML)2024.07 | — | 4.14 | 3.76 | 5.1 | 5.84 | |
| Universal-12024.07 | — | — | 3.3 | 2.3 | — | |
| Whisper Large-v22024.07 | — | 6.2 | 5.7 | 8.1 | — |