Truly unsupervised acoustic word embeddings using weak top-down constraints in encoder-decoder models

Devs

Truly unsupervised acoustic word embeddings using weak top-down constraints in encoder-decoder models | Read Paper on Bytez