Measurement of Complementarity of Recognition Systems

Combination of different speech recognition systems can be powerful technique to improve recognition performance. The success of these techniques, however, depends on the complementarity of the combined systems. In this paper, measures of complementarity of different recognition systems are proposed. These measures are based on analysis of similarity of errors made by individual systems. High correlation between these measures and actual performances of combined systems is shown in experiments, which indicates that these measures can be used to select systems suitable for combination. The measures can be computed very efficiently and they can be used even in situations where exhaustive search looking for the set of systems optimal for combination would be infeasible.

• Type: Conference
• Year: 2004
• Where: TSD
• Authors: Lukas Burget
• Keywords: Combined Systems, Recognition Systems, Signal Processing, Speech Recognition Systems, TSD 2004