Junyan Wu

weakly supervised learning (2) self-supervised learning (1) image forgery detection (1) diffusion model (1) adversarial perturbation (1) vision-language model (1) speaker identity (1) semantic contrastive learning (1) voice cloning (1) forgery detection (1) audio forgery localization (1) audio-language co-learning (1) proactive defense (1) contrastive learning (1) pixel-level localization (1)