Processed Multimodal Features for Emotion Analysis (IEMOCAP-based)

Name: Processed Multimodal Features for Emotion Analysis (IEMOCAP-based)
Creator: Nischal Mandal
Keywords: Machine Learning

Citation Author(s):: Nischal Mandal
Submitted by:: Nischal Mandal
Last updated:: Mon, 05/05/2025 - 03:03
DOI:: 10.21227/0h60-7t45

20 views

Categories:

Machine Learning

Keywords:

data mining; sentiment analysis; hate speech detection; anxiety or stress analysis; machine learning; natural language processing

ACCESS DATASET CITE

Abstract

This dataset contains preprocessed and engineered features used in the manuscript titled 'Rethinking Multimodal Sentiment Analysis: A High-Accuracy, Simplified Fusion Architecture' submitted to IEEE Transactions on Affective Computing. It includes TF-IDF vectors, audio statistics, and motion-based video features.

Instructions:

This dataset contains preprocessed and engineered features used in the manuscript titled Rethinking Multimodal Sentiment Analysis: A High-Accuracy, Simplified Fusion Architecture submitted to IEEE Transactions on Affective Computing. It includes TF-IDF vectors, audio statistics, and motion-based video features.

Dear Contributor,
I am a researcher/student working on emotion recognition and multimodal analysis in human-computer interaction. I would like to request access to the "Processed Multimodal Features for Emotion Analysis (IEMOCAP-based)" dataset to use it for academic purposes, particularly to train and evaluate models for audio-visual emotion classification. The dataset will be used strictly for non-commercial, research-focused analysis. Thank you for your support and contribution to the research community.

Benjamin IBOK Mon, 05/05/2025 - 16:27 Permalink