Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Self-adaptive Dataset Construction for Real-World Multimodal Safety Scenarios

Jingen Qu, Lijun Li, Bo Zhang, Yichen Yan, Jing Shao

TL;DR

A novel image-oriented self-adaptive dataset construction method for RMS, which starts with images and end constructing paired text and guidance responses, offering a new perspective for the construction of real-world multimodal safety datasets.

Abstract

Multimodal large language models (MLLMs) are rapidly evolving, presenting increasingly complex safety challenges. However, current dataset construction methods, which are risk-oriented, fail to cover the growing complexity of real-world multimodal safety scenarios (RMS). And due to the lack of a unified evaluation metric, their overall effectiveness remains unproven. This paper introduces a novel image-oriented self-adaptive dataset construction method for RMS, which starts with images and end constructing paired text and guidance responses. Using the image-oriented method, we automatically generate an RMS dataset comprising 35k image-text pairs with guidance responses. Additionally, we introduce a standardized safety dataset evaluation metric: fine-tuning a safety judge model and evaluating its capabilities on other safety datasets.Extensive experiments on various tasks demonstrate the effectiveness of the proposed image-oriented pipeline. The results confirm the scalability and effectiveness of the image-oriented approach, offering a new perspective for the construction of real-world multimodal safety datasets. The dataset is presented at https://huggingface.co/datasets/NewCityLetter/RMS2/tree/main.

Self-adaptive Dataset Construction for Real-World Multimodal Safety Scenarios

TL;DR

A novel image-oriented self-adaptive dataset construction method for RMS, which starts with images and end constructing paired text and guidance responses, offering a new perspective for the construction of real-world multimodal safety datasets.

Abstract

Multimodal large language models (MLLMs) are rapidly evolving, presenting increasingly complex safety challenges. However, current dataset construction methods, which are risk-oriented, fail to cover the growing complexity of real-world multimodal safety scenarios (RMS). And due to the lack of a unified evaluation metric, their overall effectiveness remains unproven. This paper introduces a novel image-oriented self-adaptive dataset construction method for RMS, which starts with images and end constructing paired text and guidance responses. Using the image-oriented method, we automatically generate an RMS dataset comprising 35k image-text pairs with guidance responses. Additionally, we introduce a standardized safety dataset evaluation metric: fine-tuning a safety judge model and evaluating its capabilities on other safety datasets.Extensive experiments on various tasks demonstrate the effectiveness of the proposed image-oriented pipeline. The results confirm the scalability and effectiveness of the image-oriented approach, offering a new perspective for the construction of real-world multimodal safety datasets. The dataset is presented at https://huggingface.co/datasets/NewCityLetter/RMS2/tree/main.

Paper Structure

This paper contains 32 sections, 11 figures, 26 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Image-oriented Construction
  4. Definition of RMS Data
  5. Safety Risk Categorization
  6. Image-oriented Construction Pipeline
  7. Pattern Generation
  8. Data Augmentation
  9. Guidance Response Generation
  10. Authenticity and Safety Review
  11. Data Analysis
  12. Data Diversity
  13. Data and Category Distribution
  14. Experiments
  15. Incremental Experiment
  16. ...and 17 more sections

Figures (11)

  • Figure 1: A conceptual sample of RMS dataset, where the image and text are safe individually and the image comes from real-world scenario.
  • Figure 2: Image-oriented method based on information complementarity.
  • Figure 3: Detailed scenarios in RMS Dataset.
  • Figure 4: The architecture of the image-oriented RMS pipeline. Starting from the real-world safe image, we generate an image-text-risk triplet, and then perform data augmentation.
  • Figure 5: The criteria of augmenting the data.
  • ...and 6 more figures