Reconciliation of human and machine speech recognition performance

• Published in: 2009 IEEE International Conference on Acoustics, Speech and Signal Processing pp. 1669 - 1672
• Main Authors: Pavel, M., Slaney, M., Hermansky, H.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.04.2009
• Subjects: Acoustic measurements; Automatic speech recognition; Availability; Biomedical engineering; Biomedical measurements; Error analysis; Error correction; Frequency; Humans; Pattern Recognition; Speech recognition
• DOI: 10.1109/ICASSP.2009.4959922
• ISBN: 9781424423538; 1424423538
• ISSN: 1520-6149

This paper focuses on resolving a number of issues that appear when the performance of human speech recognition is compared to that of automatic speech recognition. In particular human experimental data suggest that the resulting error is a product of the individual streams. On the other hand, Bayesian combination requires a multiplication of the estimates of prior probabilities and likelihoods. We show that, in principle, there is no discrepancy. The product of errors is a performance measure and human and machine performance may be consistent with this empirically established regularity. The product of probabilities is step in an algorithm to achieve the performance that may or may not be consistent with the product of errors. The main problem is that most of prior discussions failed to distinguish the performance measures from the estimates of the parameters used in the algorithm.