872
The Australian Institute of Criminology has released research on new software that can detect and match face and voice biometrics in child sexual abuse videos
An Australian Institute of Criminology study, undertaken by a team led by researchers from the University of Adelaide and San Jose State University, tested new software to detect and identify individuals in child sexual abuse videos using a combination of face and voice recognition.
- Combining face and voice biometrics improved accuracy of victim and perpetrator identification in child abuse videos, producing a true match rate of 93.8 to 98.8 percent.
- This new technology overcomes the limitation of current automated software that relies solely on face recognition, given that faces of victims and perpetrators are often hidden or obscured.
- This software may allow the analysis of large volumes of abusive videos while reducing the amount of harmful material investigators are exposed to.