Table 2 Summary of the perfect knowledge source identification approaches
Method | Cues | Methodology | Class | Media | SNs | Dataset |
|---|---|---|---|---|---|---|
[6] | PRNU | Correlation + threshold | Device | Video | YouTube | − |
[8] | PRNU-based | PCE + threshold | Device | Image video | Facebook, YouTube | VISION |
[9] | PRNU | PCE + threshold | Device | Video | YouTube | VISION |
[11] | Hybrid PRNU | PCE + threshold | Device | Image video | YouTube, Facebook | VISION |
[3] | PRNU | PCE + threshold | Device | Video | Wireless Camera | − |
[7] | PRNU | PCE + Threshold | Device | Image | Facebook, MySpace, Photobucket | − |
[12] | PRNU | PCE + threshold | Device | Image | Facebook, Google+, WhatsApp | − |
[13] | Rich features CFA CNN-based | SVM-based approaches, Open-Set Nearest Neighbors classifier | Model | Image | Flickr | FLICKR UNICAMP |
[4] | CNN-based | Dense Net (patch based) | Model | Image | Flickr | − |
[14] | RGB | CNN | Model | Image | Flickr, Yandex.Fotki, Wikipedia Commons | − |
[10] | PRNU | VGG | Device | Image | VISION |