MELD-Preprocessed / README.md
Vano04's picture
Update README.md
acd7640 verified
|
Raw History Blame Contribute Delete
618 Bytes
metadata
license: gpl-3.0
language:
  - en
pretty_name: MELD Preprocessed

MELD Preprocessed for SER

This dataset is the manually preprocessed audio only version of MELD, only audio IDs, utterance transcriptions, dialogue IDs and Utterance IDs were extracted.

S. Poria, D. Hazarika, N. Majumder, G. Naik, R. Mihalcea,
E. Cambria. MELD: A Multimodal Multi-Party Dataset
for Emotion Recognition in Conversation. (2018)
Chen, S.Y., Hsu, C.C., Kuo, C.C. and Ku, L.W.
EmotionLines: An Emotion Corpus of Multi-Party
Conversations. arXiv preprint arXiv:1802.08379 (2018).