Cultural-Aware AI Model for Emotion Recognition

Emotion AI is a research domain that aims to understand human emotions from visual or textual data. However, existing methods often ignore the influence of cultural diversity on emotional interpretation. In this paper, we propose a multimodal deep learning model that integrates cultural awareness in...

Full description

Saved in:
Bibliographic Details
Main Authors: Baradaran, Mehrdad, Zohari, Payam, Mahyar, Abtin, Motamednia, Hossein, Rahmati, Dara, Gorgin, Saeid
Format: Conference Proceeding
Language:English
Subjects:
Annotations
Deep learning
DeepFeature Extraction
Emotion recognition
Image recognition
Image Representation
Soft sensors
Speech recognition
Text Representation
Visualization
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Emotion AI is a research domain that aims to understand human emotions from visual or textual data. However, existing methods often ignore the influence of cultural diversity on emotional interpretation. In this paper, we propose a multimodal deep learning model that integrates cultural awareness into emotion recognition. Our model uses images as the primary data source and comments from individuals across different regions as the secondary data source. Our results show that our model achieves robust performance across various scenarios. Our contribution is to introduce a novel fusion approach that bridges cultural gaps and fosters a more nuanced understanding of emotions. Due to the best of our knowledge, few works are using this approach, for Emotion AI, combining different types of data sources and models. We evaluate our model on the ArtELingo dataset, which contains image-comment pairs with Chinese, Arabic, and English annotations. The experimental results in the evaluation phase demonstrate an impressive 80% recognition accuracy for the model that merges image-text features.
ISSN:2166-6784
DOI:10.1109/MVIP62238.2024.10491176