Abstract: To make deepfake detectors generalizable to different target domains, one effective way is to let the source domain for training be similar to the target domain for detection. This letter ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback