Skip to main content

Table 3 Average SimData and RealData word error rates for the REVERB Challenge evaluation set

From: Feature enhancement of reverberant speech by distribution matching and non-negative matrix factorization

  1. Comparable single and eight channel feature enhancement results from studies by Weninger et al. [6] and Tachioka et al. [30] are also presented for the most complex 1- and 8-channel back-ends. The best results for each recording condition and for both 1- and 8-channel systems are printed in boldface. Here, in addition to the abbreviations of Table 2, BF denotes beamforming in general and as an example, R1F is decoded as Room 1, Far microphone. The dashed line separates the performance-wise comparable back-ends