End-to-End Speech Translation for Code Switched Speech

Abstract

Code switching (CS) refers to the phenomenon of interchangeably using words and phrases from different languages. CS can pose significant accuracy challenges to NLP, due to the often monolingual nature of the underlying systems. In this work, we focus on CS in the context of English/Spanish conversations for the task of speech translation (ST), generating and evaluating both transcript and translation. To evaluate model performance on this task, we create a novel ST corpus derived from existing public data sets. 1 We explore various ST architectures across two dimensions: cascaded (transcribe then translate) vs end-toend (jointly transcribe and translate) and unidirectional (source → target) vs bidirectional (source ↔ target). We show that our ST architectures, and especially our bidirectional end-to-end architecture, perform well on CS speech, even when no CS training data is used.

Code-Switching without Switching: Language Agnostic End-to-End Speech Translation

“…Most similar to our work is the model E2E BIDIRECT SHARED of [16, figure 3G]. However, the difference to our work is, that [16] uses CS data, where they need transcriptions and translations, as well as annotations which words are from which language, and they focus on intra-sentential CS. Furthermore, they first generate a transcription and therefore have to explicitly detect which language is spoken in each part of the audio.…”