The online sharing and viewing of Child Sexual Abuse Material (CSAM) are growing fast, such that human experts can no longer handle the manual inspection. However, the automatic classification of CSAM is a challenging field of research, largely due to the inaccessibility of target data that is - and should forever be - private and in sole possession of law enforcement agencies. To aid researchers in drawing insights from unseen data and safely providing further understanding of CSAM images, we propose an analysis template that goes beyond the statistics of the dataset and respective labels. It focuses on the extraction of automatic signals, provided both by pre-trained machine learning models, e.g., object categories and pornography detection, as well as image metrics such as luminance and sharpness. Only aggregated statistics of sparse signals are provided to guarantee the anonymity of children and adolescents victimized. The pipeline allows filtering the data by applying thresholds to each specified signal and provides the distribution of such signals within the subset, correlations between signals, as well as a bias evaluation. We demonstrated our proposal on the Region-based annotated Child Pornography Dataset (RCPD), one of the few CSAM benchmarks in the literature, composed of over 2000 samples among regular and CSAM images, produced in partnership with Brazil's Federal Police. Although noisy and limited in several senses, we argue that automatic signals can highlight important aspects of the overall distribution of data, which is valuable for databases that can not be disclosed. Our goal is to safely publicize the characteristics of CSAM datasets, encouraging researchers to join the field and perhaps other institutions to provide similar reports on their benchmarks.
Human-Centric Perception for Child Sexual Abuse Imagery
Universidade Federal de Minas Gerais (UFMG)(米纳斯吉拉斯联邦大学);Federal Police of Brazil(巴西联邦警察局);Universidade Estadual de Campinas (UNICAMP)(坎皮纳斯州立大学);University of Sheffield(谢菲尔德大学)
CSA-Graphs: A Privacy-Preserving Structural Dataset for Child Sexual Abuse Research
Universidade Estadual de Campinas (UNICAMP)(坎皮纳斯州立大学);Instituto Federal de Educação, Ciência e Tecnologia de Minas Gerais (IFMG)(米纳斯吉拉斯联邦教育、科学和技术研究所);Universidade de São Paulo (USP)(圣保罗大学);Universidade Federal de Minas Gerais (UFMG)(米纳斯吉拉斯联邦大学);Polícia Federal (PF)(联邦警察);University of Sheffield(谢菲尔德大学)
Leveraging Self-Supervised Learning for Scene Classification in Child Sexual Abuse Imagery
Instituto de Computação, Universidade Estadual de Campinas (UNICAMP)(计算学院,坎皮纳斯州立大学(UNICAMP));Departamento de Ciência da Computação, Universidade Federal de Minas Gerais (UFMG)(计算机科学系,巴西矿务联邦大学(UFMG));Departamento de Polícia Federal (DPF)(联邦警察部门);Department of Computer Science, University of Sheffield(计算机科学系,谢菲尔德大学)
Classifying by Proxy: Explainable and Reproducible Ensemble of Proxy Tasks for Child Sexual Abuse Imagery Classification
Instituto de Ciências Matemáticas e de Computação (ICMC), Universidade Estadual de São Paulo (USP)(圣保罗州立大学数学与计算机科学学院);Instituto de Computação (IC), Universidade Estadual de Campinas (UNICAMP)(坎皮纳斯州立大学计算机学院);Departamento de Ciência da Computação, Universidade Federal de Minas Gerais (UFMG)(巴西矿务大学计算机科学系);Instituto Federal de Educação, Ciência e Tecnologia de Minas Gerais (IFMG)(米纳斯吉拉斯州联邦教育、科学和技术研究院)
Minimizing Risk Through Minimizing Model-Data Interaction: A Protocol For Relying on Proxy Tasks When Designing Child Sexual Abuse Imagery Detection Models
Department of Computer Science, University of Sheffield(谢菲尔德大学计算机科学系)
The Nipple-Areola Complex for Criminal Identification
School of Computer Science and Engineering, Nanyang Technological University(南洋理工大学计算机科学与工程学院);Faculty of Medicine, Medical University, Wroclaw(弗罗茨瓦夫医科大学医学院);The MITRE Corporation(MITRE公司)