Results 1 - 10
of
2,364
Table 2: Performances WER [%] of MFCC-based acoustic models trained on artificially reverberated speech for various reverberant testing conditions.
"... In PAGE 3: ... The wall absorp- tion coefficients of the reverberant room simulator were chosen to get specific reverberation times. Table2 reports cross-testing results.... ..."
Table 3. Data sets for the recognizers trained with Verbmobil data ( mic. dist. = microphone distance, CT art. rev. = close-talking artificially reverberated)
2005
Cited by 2
Table 4. Word error rates in percent for the baseline SRI recognizer without mean subtraction. Results without/with speaker adaptation are before/after the slashes. The test sets were artificially reverberated TIDIGITS and tabletop-mic Meeting Recorder Digits.
2001
Cited by 3
Table 1 Reverberation times of three conditions used in the experiments Reverberation
in in
"... In PAGE 8: ...3. Processed speech stimuli Reverberation times ( Table1 ) were measured in three auditoria in Japan: Hamming Hall in Higa- shi-yamato City (Rev A), Hamming Hall in Higa- shi-yamato City without reflection boards (Rev B), and Strings booth in the NHK Hall (Rev C). The early decay time (EDT), which was used to extrap- olate reverberation time by measuring the required time for the first 10 dB down of the decay curve, was employed for measuring reverberation time (Jordan, 1980).... ..."
Tables. INTRODUCTION
Table 10: Word error rates - reverberant speech
1999
Cited by 52
Table of contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2001
Cited by 61
Results 1 - 10
of
2,364