LIUM’s SMT Machine Translation Systems for WMT 2012

By Christophe Servan, Patrik Lambert, Anthony Rousseau, Holger Schwenk, Loïc Barrault. 2012

16 citations recorded by the source corpus.

Citations in the corpus

• Recurrent Neural Network Regularization (2014, neutral): “While our LSTM does not beat the phrase-based LIUM SMT system (Schwenk et al., 2011), our results show that dropout improves the translation performance of the LSTM.”