Deep neural networks for oral and written language processing (RAPACE)

Date: 10/2017 - 09/2020
Funding: RFI AltanStic 2020
Call: Défis scientifiques 2016
Partners: LINA (France)

The RAPACE project focuses on the implementation of a deep neural architecture, associating specialized hidden layers, to develop a fully neural system of speech processing. The aim is to develop a system for detecting named entities in an audio stream that will rely solely on a deep neural network, as opposed to what has been done for nearly two decades using successive processing chains.