Table 3 Voice-tag word accuracy obtained by sequential voice-to-phoneme conversion with bigram phoneme histogram.
From: Voice-to-Phoneme Conversion Algorithms for Voice-Tag Applications in Embedded Platforms
| Â | Word accuracy (%) with | |||
|---|---|---|---|---|
| Â | bigram phoneme histogram | |||
Sequentially combine hypotheses from | Single sample | Two samples | Three samples | Baseline |
CI HMMs | 87.7 | 89.9 | 90.4 | 91.67 |
CD HMMs | 87.3 | 94.0 | 95.2 | 92.69 |