A Deep Network for Arousal-Valence Emotion Prediction with Acoustic-Visual Cues

TL;DR

This paper presents a deep learning approach that integrates acoustic and visual cues to predict arousal and valence emotions, aiming to improve emotion recognition accuracy.

Contribution

It introduces a novel deep network architecture specifically designed for multimodal emotion prediction using acoustic and visual data.

Findings

Achieved competitive results in the 2018 Emotion Behavior Challenge.

Demonstrated the effectiveness of multimodal cues in emotion prediction.

Provided a detailed methodology for emotion recognition using deep learning.

Abstract

In this paper, we comprehensively describe the methodology of our submissions to the One-Minute Gradual-Emotion Behavior Challenge 2018.