Processed Multimodal Feature Datasets for Emotion Recognition Using IEMOCAP and MELD

This repository contains the minimal processed datasets used in the study on multimodal emotion recognition. It includes pre-processed multimodal feature files derived from the two benchmark datasets named: IEMOCAP and MELD. These files were generated using the authors’ publicly available preprocessing pipeline and were used for training and evaluating the proposed deep learning model. Due to licensing restrictions, the original raw datasets are not included. These files are sufficient to reproduce the experimental results reported in the associated manuscript.