MAT-MRNN-STD
Unsupervised Spoken Term Detection with Multi-target sequence to sequence autoendocer
Acoustic Tokenizer is modified from zero-resource-speech-tokenizer(zrst)
Required Packages
- HTK
- SRILM
Unsupervised Spoken Term Detection with Multi-target sequence to sequence autoendocer
Acoustic Tokenizer is modified from zero-resource-speech-tokenizer(zrst)