The Sensitive Media Analysis (SMA) project aims at researching solutions to combine different and complementary data representations and pattern classifiers for detecting sensitive content in digital images and videos.
Sensitive media can be defined as the digital content whose depiction to particular audiences (e.g., children or unwary spectators), at particular places (e.g., at work, at school, in the church) may inflict harm (e.g., trauma, shock, or fear) due to its inappropriateness. Typical representatives include – but are not limited to – scenes depicting pornography and violence, animal cruelty and child abuse, hate speech, etc.
The innovation aspects of the project reside on the development of solutions that are amenable to deployment on mobile devices (e.g., smartphones and tablets), observing their constraints of memory footprint, processing power, and runtime responsiveness.
- Prof. Anderson Rocha (PI)
- Prof. Siome Goldenstein (PI)
- Prof. Eduardo Valle (PI)
- Dr. Vanessa Testoni (Samsung Collaborator)
- Dr. Sandra Avila (Postdoc)
- Daniel Moreira (PhD Student)
- Mauricio Perez (MSc Student)
- Daniel Moraes (Developer)