Table 1 Summary of recording conditions and data set parameters. SimData denotes artificially reverberated speech data with real RIRs and RealData denotes true recordings made in a reverberant room
Recording type | Room size | T 60 (s) | Near mic. | Far mic. |
|---|---|---|---|---|
distance (m) | distance (m) | |||
SimData | Small | 0.25 | 0.5 | 2.0 |
SimData | Medium | 0.5 | 0.5 | 2.0 |
SimData | Large | 0.7 | 0.5 | 2.0 |
RealData | Large | 0.7 | 1.0 | 2.5 |
Data set | Recording | Number of | Utterances | Duration (h) |
type | speakers | |||
Training | Clean speech | 92 | 7961 | 17.5 |
MC training | Similar to | 92 | 7961 | 17.5 |
SimData | ||||
Development | SimData | 20 | 1484 | 3.3 |
RealData | 5 | 179 | 0.3 | |
Evaluation | SimData | 20 | 2176 | 4.8 |
RealData | 10 | 372 | 0.6 |