ICM-Assistant: Instruction-tuning Multimodal Large Language Models for Rule-based Explainable Image Content Moderation

TL;DR

This paper introduces ICM-Assistant, a rule-based multimodal large language model designed for explainable and accurate image content moderation, outperforming existing models in classification and explanation quality.

Contribution

It presents a novel rule-based dataset generation pipeline and a specialized model that enhances flexibility, explainability, and accuracy in image content moderation tasks.

Findings

Significant improvement in moderation classification accuracy (36.8%)

Enhanced explanation quality (26.6%) over existing models

Effective application of rule-based prompts in multimodal models

Abstract

Controversial contents largely inundate the Internet, infringing various cultural norms and child protection standards. Traditional Image Content Moderation (ICM) models fall short in producing precise moderation decisions for diverse standards, while recent multimodal large language models (MLLMs), when adopted to general rule-based ICM, often produce classification and explanation results that are inconsistent with human moderators. Aiming at flexible, explainable, and accurate ICM, we design a novel rule-based dataset generation pipeline, decomposing concise human-defined rules and leveraging well-designed multi-stage prompts to enrich short explicit image annotations. Our ICM-Instruct dataset includes detailed moderation explanation and moderation Q-A pairs. Built upon it, we create our ICM-Assistant model in the framework of rule-based ICM, making it readily applicable in real practice. Our ICM-Assistant model demonstrates exceptional performance and flexibility. Specifically, it significantly outperforms existing approaches on various sources, improving both the moderation classification (36.8% on average) and moderation explanation quality (26.6% on average) consistently over existing MLLMs. Code/Data is available at https://github.com/zhaoyuzhi/ICM-Assistant.