CSA-Graphs: A Privacy-Preserving Structural Dataset for Child Sexual Abuse Research

TL;DR

CSA-Graphs is a novel privacy-preserving dataset using structural graph representations to facilitate child sexual abuse imagery research without sharing explicit visual content.

Contribution

Introduces CSA-Graphs, a structural dataset with scene and skeleton graphs that maintain classification utility while ensuring privacy and legal compliance.

Findings

Both graph modalities retain classification-relevant information.

Combining scene and skeleton graphs improves performance.

Enables research under strict privacy constraints.

Abstract

Child Sexual Abuse Imagery (CSAI) classification is an important yet challenging problem for computer vision research due to the strict legal and ethical restrictions that prevent the public sharing of CSAI datasets. This limitation hinders reproducibility and slows progress in developing automated methods. In this work, we introduce CSA-Graphs, a privacy-preserving structural dataset. Instead of releasing the original images, we provide structural representations that remove explicit visual content while preserving contextual information. CSA-Graphs includes two complementary graph-based modalities: scene graphs describing object relationships and skeleton graphs encoding human pose. Experiments show that both representations retain useful information for classifying CSAI, and that combining them further improves performance. This dataset enables broader research on computer vision methods for child safety while respecting legal and ethical constraints.