1 paper · 1 filter
Hannah Lee, Changyeon Lee, Kevin Farhat +4
Multimodal generative models are rapidly evolving, leading to a surge in the generation of realistic video and audio that offers exciting possibilities but also serious risks. Deep…