Automatically augmenting an emotion dataset improves classification using audio.

Egor Lakomkin, Cornelius Weber, Stefan Wermter

Browse the full EACL paper archive.