Weekly Report:
1, Try to add language identification information to bi-lingual asr, no good performance observed; 2, Test the rnn-dae on AURORA4 dataset.
Target of next week:
1, Continue the investigation of bi-lingual speech recognition problem.