A database of German emotional speech

TLDR

The database was created by having ten actors (five female, five male) simulate seven emotions in 10 short and 10 long German utterances, recorded in an anechoic chamber with high‑quality microphones and electroglottography, yielding roughly 800 sentences that were later evaluated in a perception test. Utterances recognized with over 80 % accuracy and rated natural by more than 60 % of listeners were phonetically transcribed with detailed markers, and the resulting database is publicly available at http://www.expressive‑speech.net/emodb/.

Abstract

The article describes a database of emotional speech. Ten actors (5 female and 5 male) simulated the emotions, producing 10 German utterances (5 short and 5 longer sentences) which could be used in everyday communication and are interpretable in all applied emotions. The recordings were taken in an anechoic chamber with high-quality recording equipment. In addition to the sound electro-glottograms were recorded. The speech material comprises about 800 sentences (seven emotions * ten actors * ten sentences + some second versions). The complete database was evaluated in a perception test regarding the recognisability of emotions and their naturalness. Utterances recognised better than 80% and judged as natural by more than 60% of the listeners were phonetically labelled in a narrow transcription with special markers for voice-quality, phonatory and articulatory settings and articulatory features. The database can be accessed by the public via the internet (http://www.expressive-speech.net/emodb/).