Emotions are related to multi-faceted clues, such as facial expressions, prosody, gestures (or micro-gestures), etc. It is inappropriate to identify emotions from just one aspect. EMER provides a general format for all emotion-related tasks, aiming to integrate multiple clues and generate more comprehensive descriptions.
Details can be found in our paper: Explainable Multimodal Emotion Reasoning
@article{lian2023explainable,
title={Explainable Multimodal Emotion Reasoning},
author={Lian, Zheng and Sun, Licai and Xu, Mingyu and Sun, Haiyang and Xu, Ke and Wen, Zhuofan and Chen, Shun and Liu, Bin and Tao, Jianhua},
journal={arXiv preprint arXiv:2306.15401},
year={2023}
}To construct the initial dataset, we select samples from MER 2023.
(1) Download Raw MER 2023 Dataset
To download the dataset, please fill out an EULA and send it to [email protected].
(2) EMER-V1
Due to the high annotation cost, we only select 100 non-neutral samples to form the initial dataset. See https://arxiv.org/abs/2306.15401v3 for more details. We provide the annotated results in ./EMER/dataset-v1. (100 samples)
(3) EMER-V2
However, the description obtained in previous manner is short and cannot cover multi-faceted clues. Therefore, we use GPT-4V to provide initial annotations, combining with manual check and ChatGPT's reasoning capabilities. See https://arxiv.org/abs/2306.15401 for more details. We provide the annotated results in ./EMER/dataset-v2. (332 samples)
We aim to train an audio-video-text aligned model to deal with explainable multimodal emotion reasoning. Specifically, we modify the code in Video-LLaMA and make it support for audio-video-text aligned training. Meanwhile, we unify the input for different datasets, remove some unnecessary modules, add model.eval() during inference, etc. More details can be found in our implementation. We provide the code in ./AffectGPT. Currently, we only provide the code training with EMER-V1 dataset.
More experimental results can be found in https://arxiv.org/abs/2306.15401v3.