Table 4 System overall performance with different configurations with respect to native origin of speakers and using the Arabic LDC-WestPoint corpus.
From: Experiments on Automatic Recognition of Nonnative Arabic Speech
N/N Exp. 1.a | N/NN Exp. 2.a | NN/N Exp. 3.a | NN/NN Exp. 4.a | M/M Exp. 5.a | N/N Exp. 1.b | N/NN Exp. 2.b | NN/N Exp. 3.b | NN/NN Exp. 4.b | M/M Exp. 5.b | |
|---|---|---|---|---|---|---|---|---|---|---|
Corr (%) | 52,01% | 46,44% | 43,03% |
| 50,32% |
| 89,46% | 93,98% | 96,78% | 98,85% |
Del (%) | 2,33% | 2,24% | 2,43% | 2,74% | 2,37% | 0,00% | 0,00% | 0,01% | 0,00% | 0,00% |
Sub (%) | 45,66% | 51,32% | 54,54% | 4,27% | 47,31% | 0,01% | 10,22% | 5,08% | 2,89% | 0,01% |
Without bigram-based language model | With bigram-based language model | |||||||||
