Multimodal Adversarial Defense for Vision-Language Models by Leveraging One-To-Many Relationships.

Futa Waseda, Antonio Tejero-de-Pablos, Isao Echizen

Browse the full WACV paper archive.