Introduction

This is a pytorch implementation of "Brain-Inspired Distributed Fusion for Robust Multimodal Sentiment Analysis"

Requirements

• pytorch == 2.6.0 + 12.7
• transformers == 4.44.2
• python == 3.12.3

How to Run

• Download dataset from https://github.com/thuiar/MMSA
• In "config\config_regression.py" set the file path to the path where your dataset files are located.
• python run.py -d mosi -g 0

evaluation

• The evaluation metrics are the same as LNLN.
• python robust_evaluation.py -d mosi -g 0

Acknowledge

We begin our work on the basis of SELF-MM and LNLN initially, so the whole code architecture is similar to them, including the data process, data loader and evaluation metrics. Thanks to their open source spirit for saving us a lot of time.