Neural Discrete Representation Learning

Explore this paper's citation graph

Summary

Pairing these representations with an autoregressive prior, the model can generate high quality images, videos, and speech as well as doing high quality speaker conversion and unsupervised learning of phonemes, providing further evidence of the utility of the learnt representations.

Type
preprint
Published
2017-11-02
Cited by
7,851
References
43
Access
Open access

Keywords

Autoencoder, Computer science, Autoregressive model, Representation (politics), Key (lock)

References

Cited by

Related papers