Child Automatic Speech Recognition on MyST 10h
11.57WERWavLM + ΔHuBERT
Evaluation Results
| Method | Links | |
|---|---|---|
| WavLM + ΔHuBERTFusion Method=Concat2026.01 | 11.57 | |
| WavLM + ΔW2V2Fusion Method=Concat2026.01 | 11.61 | |
| WavLMModel Type=Single fine-tuned SSL2026.01 | 11.95 | |
| HuBERTModel Type=Single fine-tuned SSL2026.01 | 12.95 | |
| W2V2Model Type=Single fine-tuned SSL2026.01 | 13.47 |