MAIL_ACL24

Modality-Aware Integration with Large Language Models for Knowledge-based Visual Question Answering

Pre-processing

In this part, we:

• Transform images into captions then Scene Graphs;
• Extract the topic entities in the questions;
• Construct the Concept Graphs based on previous outputs.

Localization of MiniGPT-4

Please check the implementation details in Localized MiniGPT-4.