AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Quantum-inspired Neural Network for Conversational Emotion Recognition

Qiuchi Li, Dimitris Gkoumas, Alessandro Sordoni, Jian-Yun Nie, Massimo Melucci

AAAI 2021 · Architectures and efficiency › RNNs and normalization · approach: Associative memory networks · 55 citations (OpenAlex)

Read the paper

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use QMNN instead of DialogueRNNIEMOCAP emotion recognition · IEMOCAPaccuracy: 62.42 -> 60.84Hurt
Use QMNN instead of DialogueRNNIEMOCAP emotion recognition · IEMOCAPF1: 60.48 -> 59.88Hurt
Use QMNN instead of the strongest baselineMELD emotion recognition · MELDaccuracy: 60.54 -> 60.81Improved
Use QMNN instead of the strongest baselineMELD emotion recognition · MELDF1: 57.79 -> 58.00Improved
Add quantum mixture fusion rather than concatenating unimodal features (ablation)QMNN · MELDF1: 0.5623 -> 0.5800Improved
Add phase encoding rather than ignoring utterance phases (ablation)QMNN · MELDF1: 0.5716 -> 0.5800Improved
Add recurrent contextual modeling (ablation)QMNN · MELDF1: 0.5684 -> 0.5800Improved
Use QRNN rather than a classical GRU (ablation)QMNN · MELDF1: 0.5686 -> 0.5800Improved
Use orthogonal quantum measurement rather than semantic measurement (ablation)QMNN · MELDF1: 0.5673 -> 0.5800Improved
Use quantum measurement rather than a fully connected layer on flattened density matrices (ablation)QMNN · MELDF1: 0.5700 -> 0.5800Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy