Conversational Memory Network for Emotion Recognition in Dyadic Dialogue Videos

@inproceedings{Hazarika2018ConversationalMN,
  title={Conversational Memory Network for Emotion Recognition in Dyadic Dialogue Videos},
  author={Devamanyu Hazarika and Soujanya Poria and Amir Zadeh and Erik Cambria and Louis-Philippe Morency and Roger Zimmermann},
  booktitle={NAACL-HLT},
  year={2018}
}
Conversational Memory Network: To classify emotion of utterance ui, corresponding histories (hista and histb) are taken. Each history, histλ, contains the preceding K utterances by person Pλ. Histories are modeled into memories and utilized as follows, Memory Representation: Memory representation Mλ = [mλ, ...,mλ ] for histλ is generated using a GRU, λ ∈ {a, b}. Memory Input: Attention mechanism is used to read Mλ. Relevance of each memory mλ’s context with ui is computed using a match… CONTINUE READING

Figures, Tables, and Topics from this paper.

Explore Further: Topics Discussed in This Paper

Similar Papers

Citations

Publications citing this paper.
SHOWING 1-10 OF 15 CITATIONS

DialogueRNN: An Attentive RNN for Emotion Detection in Conversations

VIEW 10 EXCERPTS
CITES METHODS & BACKGROUND
HIGHLY INFLUENCED

An Interaction-aware Attention Network for Speech Emotion Recognition in Spoken Dialogs

  • ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
  • 2019

References

Publications referenced by this paper.
SHOWING 1-3 OF 3 REFERENCES

Ensemble of SVM trees for multimodal emotion recognition

  • Proceedings of The 2012 Asia Pacific Signal and Information Processing Association Annual Summit and Conference
  • 2012
VIEW 1 EXCERPT

Conversational Memory Network for Emotion Recognition in Dyadic Dialogue Videos

Viktor Rozgic, Saleem Sankaranarayanan Ananthakrishnan, Rohit Kumar, Rohit Prasad