Acoustic model meets Language model 🤔

1. CTC

Code

• ctc-decode: https://gitlab.com/sburud/master/-/tree/master/speechlm/ctc-decode

References

Streaming Small Footprint Keyword Spotting Using Seq2Seq Models

• https://arxiv.org/pdf/1710.09617.pdf

Unrestricted Vocabulary Keyword Spotting using LSTM-CTC

• https://www.isca-speech.org/archive_v0/Interspeech_2016/pdfs/0753.PDF