N-Gram and RNN-LM Language Model Integration for End-to-End Amazigh Speech Recognition

Today's article comes from the journal of Machine Learning and Knowledge Extraction. The authors are Telmem et al., from Moulay Ismail University, in Morocco. In this paper, they're trying to guide the outputs of a language model, so that it can translate a language for which there is very little training data.

DOI: 10.3390/make7040164