1. Isolated speech commands recognition by using hybrid approach
  2. Speaker identification accuracy improvement using BLSTM neural network