Learning Strategies for Sensitive Content Detection

• DOI: 10.3390/electronics12112496
• Publication Date: 2023-06-02T05:33:54Z
• Authors: Daniel Povedano Á...; Ana Lucila Sandov...; Javier Portela Ga...; Luis Javier Garcí...
• Keywords: Pornography; Blocking (statistics)

Currently, the volume of sensitive content on Internet, such as pornography and child pornography, amount time that people spend online (especially children) have led to an increase in distribution (e.g., images children being sexually abused, real-time videos abuse, grooming activities, etc.). It is therefore essential effective IT tools automate detection blocking this type material, manual filtering huge volumes data practically impossible. The goal study carry out a comprehensive review different learning strategies for available literature, from most conventional techniques cutting-edge deep algorithms, highlighting strengths weaknesses each, well datasets used. performance scalability proposed work depend heterogeneity dataset, feature extraction (hashes, visual, audio, etc.) algorithms. Finally, new lines research sensitive-content are presented.