Multilingual acoustic models for the recognition of non-native speech | IEEE Conference Publication | IEEE Xplore