Skip to main content

Table 6 WER results vs. the length of the local window

From: Advanced recurrent network-based hybrid acoustic models for low resource speech recognition

Model the length of local window  
  10 20 30 40 50
LW-BLSTM-fbank 48.8 48.3 48.3 48.3 48.3
LW-BLSTM-MBN 44.5 43.9 43.9 43.9 43.8
\