This paper introduces AbstractNet, a generative model for high density inputs. The model suggests a method that uses unsupervised learning to generate feature maps. The model drastically improves the performances of raw audio generation by reducing the required amount of input data and computing power necessary to achieve a similar result when compared to the state of the art.
CITATION STYLE
Musarais, B. (2018). Abstractnet: a generative model for high density inputs. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10710 LNCS, pp. 462–469). Springer Verlag. https://doi.org/10.1007/978-3-319-72926-8_38
Mendeley helps you to discover research relevant for your work.