Pertsch: A Corpus of Persian and German Based on Different Speech Elicitation Tasks

Abstract

The Pertsch Corpus is a bilingual speech corpus designed to investigate speech production across different communicative situations in Persian and German. The corpus contains recordings from 60 adult speakers (30 Persian, 30 German; balanced for gender), each completing a fixed sequence of seven speech elicitation tasks ranging from controlled read speech to more openended communicative conditions. The elicitation tasks include: • formal read speech • informal read speech • spontaneous conversation • picture description • autobiographical storytelling • informal voice messages • formal voice messages