Computational and Mathematical Methods in Medicine / 2013 / Article / Tab 4

Research Article

Estimation of Phoneme-Specific HMM Topologies for the Automatic Recognition of Dysarthric Speech

Table 4

Recognition accuracies obtained with different training schemes for the baseline SD ASR systems across all dysarthric speakers.

SpeakerTraining scheme 1
SD Bakis-1SD Bakis-2SD ErgodicAverage scheme 1

FB77.7372.9376.4275.69
MH73.8069.4373.3672.20
BB75.6963.3069.2769.42
LL69.2067.8666.9668.01
JF60.8948.0055.5654.82
RL50.6748.4443.5647.56
RK44.3015.3525.4428.36
BK13.276.6426.9915.63
BV48.6450.9143.6447.73
SC44.5519.5527.7330.61

SpeakerTraining scheme 2
SD Bakis-1SD Bakis-2SD ErgodicAverage scheme 2

FB72.4963.7665.9467.40
MH66.3858.9557.6460.99
BB69.7254.1362.3962.08
LL64.7357.1462.0561.31
JF49.7843.1152.0048.30
RL51.1142.6744.8946.22
RK38.1639.0439.4738.89
BK31.4218.5836.2828.76
BV55.9153.1847.7352.27
SC44.0943.1848.1845.15

We are committed to sharing findings related to COVID-19 as quickly as possible. We will be providing unlimited waivers of publication charges for accepted research articles as well as case reports and case series related to COVID-19. Review articles are excluded from this waiver policy. Sign up here as a reviewer to help fast-track new submissions.