AV-EmoDialog uses speech and face encoders with a large language model to generate emotion-aware dialogue responses from audio-visual input, reporting better emotional alignment than the compared baselines.
Non-verbal communication in human social interaction
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
AV-EmoDialog: Chat with Audio-Visual Users Leveraging Emotional Cues
AV-EmoDialog uses speech and face encoders with a large language model to generate emotion-aware dialogue responses from audio-visual input, reporting better emotional alignment than the compared baselines.